Produced by Chris Curnow, Paul Clark and the Online
Distributed Proofreading Team at http://www.pgdp.net (This
file was produced from images generously made available
by The Internet Archive)









THE EARLY NORMAN CASTLES OF THE BRITISH ISLES


[Illustration: MOTTE-CASTLES FROM THE BAYEUX TAPESTRY.]




                       THE EARLY NORMAN CASTLES
                         OF THE BRITISH ISLES

                          BY ELLA S. ARMITAGE

       HONORARY FELLOW OF THE SOCIETY OF ANTIQUARIES OF SCOTLAND

   AUTHOR OF "THE CHILDHOOD OF THE ENGLISH NATION"; "THE CONNECTION
  OF ENGLAND AND SCOTLAND"; "AN INTRODUCTION TO ENGLISH ANTIQUITIES,"
                              ETC., ETC.

                WITH PLANS BY D. H. MONTGOMERIE, F.S.A.

                                LONDON
                   JOHN MURRAY, ALBEMARLE STREET, W.
                                 1912




ERRATA


Page 34, _note_ 1.--_For_ "construerat" _read_ "construxerat."

Page 40, line 9.--_For_ "there was only one motte, the site of the
castle of the Norman Giffards is now almost obliterated," _read_ "there
was only one motte, site of the castle of the Norman Giffards, now
almost obliterated."

Page 133, line 16.--_For_ "1282" _read_ "1182."

Page 145, _note_ 1.--_For_ "Legercestria" _read_ "Legecestria."

Page 147, line 15.--Delete comma after "castle."

Page 216, _note_ 2.--_For_ "instalment" _read_ "statement."

Page 304, _note_ 3.--_For_ "Galloway, Wigton, Kirkcudbright, and
Dumfries," _read_ "Galloway (Wigton, Kirkcudbright, and Dumfries)."




PREFACE


Some portions of this book have already appeared in print. Of these,
the most important is the _catalogue raisonné_ of early Norman
castles in England which will be found in Chapter VII., and which was
originally published in the _English Historical Review_ (vol. xix.,
1904). It has, however, been enlarged by the inclusion of five fresh
castles, and by notes upon thirty-four others, of which the article in
the _Review_ gave only the names; the historical notes in that essay
being confined to the castles mentioned in Domesday Book.

The chapter on Irish mottes appeared in the _Antiquary_ (vol. xlii.,
1906), but it has been revised, corrected, and added to. Portions of a
still earlier paper, read before the Society of Antiquaries of Scotland
in March 1900, are incorporated in various parts of the book, but these
have been recast in the fuller treatment of the subject which is aimed
at here.

The rest of the work is entirely new. No serious attempt had been made
to ascertain the exact nature of Saxon and Danish fortifications by a
comparison of the existing remains with the historical records which
have come down to us, until the publication of Mr Allcroft's valuable
book on _Earthwork of England_. The chapters on Saxon and Danish
earthworks in the present volume were written before the appearance of
his book, though the results arrived at are only slightly different.

In Chapter V. an effort is made to trace the first appearance of
the private castle in European history. The private castle is an
institution which is often carelessly supposed to have existed from
time immemorial. The writer contends that it only appears after the
establishment of the feudal system.

The favourable reception given by archæologists to the paper read
before the Scottish Society led the writer to follow up this
interesting subject, and to make a closer study of the motte-castles
of Wales, Scotland, and Ireland. The book now offered is the fruit
of eleven years of further research. The result of the inquiry is
to establish the theory advanced in that earlier paper, that these
castles, in the British Islands, are in every case of Norman origin.

The writer does not claim to have originated this theory. Dr Round was
the first to attack (in the _Quarterly Review_, 1894) the assertion of
the late Mr G. T. Clark that the moated mound was a Saxon castle. Mr
George Neilson continued the same line of argument in his illuminating
paper on "The Motes in Norman Scotland" (_Scottish Review_, vol.
xxxii., 1898).[1] All that the writer claims is to have carried the
contention a stage further, and to have shown that the private castle
did not exist at all in Britain until it was brought here by the
Normans.

The author feels that some apology is necessary for the enormous length
of Chapter VII., containing the catalogue of Early English castles. It
may be urged in extenuation that much of the information it contains
has never before appeared in print, seeing that it has been taken from
unpublished portions of the Pipe Rolls; further, that contemporary
authorities have in all cases been used, and that the chapter contains
a mass of material, previously scattered and almost inaccessible, which
is here for the first time collated, and placed, as the author thinks,
in its right setting. It is hoped that the chapter will prove a useful
storehouse to those who are working at the history of any particular
castle mentioned in the list.

To many it may seem a waste of labour to devote a whole book to the
establishment of a proposition which is now generally adopted by the
best English archæologists; but the subject is an important one, and
there is no book which deals with it in detail, and in the light of the
evidence which has recently been accumulated. The writer hopes that
such fuller statement of the case as is here attempted may help not
only to a right ascription of British castle-mounds, and of the stone
castles built upon many of them, but may also furnish material to the
historian who seeks to trace the progress of the Norman occupation.

Students of the architecture of castles are aware that this subject
presents much more difficult questions than does the architecture
of churches. Those who are seriously working on castle architecture
are very few in number, and are as yet little known to the world at
large. From time to time, books on castles are issued from the press,
which show that the writers have not even an idea of the preliminary
studies without which their work has no value at all. It is hoped that
the sketch of castle architecture from the 10th century to the 13th,
which is given in the last chapter, may prove a useful contribution to
the subject, at any rate in its lists of dated castles. The Pipe Rolls
have been too little used hitherto for the general history of castle
architecture, and no list has ever been published before of the keeps
built by Henry II. But without the evidence of the Pipe Rolls we are
in the land of guesswork, unsupported, as a rule, by the decorative
details which render it easy to read the structural history of most
churches.

My warmest thanks are due to Mr Duncan H. Montgomerie, F.S.A., for
his generous labour on the plans and illustrations of this book, and
for effective assistance in the course of the work, especially in
many toilsome pilgrimages for the purpose of comparing the Ordnance
Survey with the actual remains. I also owe grateful thanks to Mr
Goddard H. Orpen, R.I.A., for most kindly revising the chapter on
Irish mottes; to Mr W. St John Hope (late Assistant Secretary of the
Society of Antiquaries), for information on many difficult points;
to Mr Harold Sands, F.S.A., whose readiness to lay his great stores
of knowledge at my disposal has been always unfailing; to Mr George
Neilson, F.S.A.Scot., for most valuable help towards my chapter on
Scottish mottes; to Mr Charles Dawson, F.S.A., for granting the use
of his admirable photographs from the Bayeux Tapestry; to Mr Cooper,
author of the _History of York Castle_, for important facts and
documents relating to his subject; to the Rev. Herbert White, M.A.,
and to Mr Basil Stallybrass, for reports of visits to castles; and to
correspondents too numerous to mention who have kindly, and often very
fully, answered my inquiries.

                                                      ELLA S. ARMITAGE.

 WESTHOLM,
 RAWDON, LEEDS.




CONTENTS


                                                          PAGE

 PREFACE                                                   vii


 CHAPTER I

 INTRODUCTORY                                                1


 CHAPTER II

 ANGLO-SAXON FORTIFICATIONS                                 11


 CHAPTER III

 ANGLO-SAXON FORTIFICATIONS--CONTINUED                      31


 CHAPTER IV

 DANISH FORTIFICATIONS                                      48


 CHAPTER V

 THE ORIGIN OF PRIVATE CASTLES                              63


 CHAPTER VI

 DISTRIBUTION AND CHARACTERISTICS OF MOTTE-CASTLES          80


 CHAPTER VII

 THE CASTLES OF THE NORMANS IN ENGLAND                      94


 CHAPTER VIII

 MOTTE-CASTLES IN NORTH WALES                              251


 CHAPTER IX

 MOTTE-CASTLES IN SOUTH WALES                              273


 CHAPTER X

 MOTTE-CASTLES IN SCOTLAND                                 302


 CHAPTER XI

 MOTTE-CASTLES IN IRELAND                                  323


 CHAPTER XII

 STONE CASTLES OF THE NORMAN PERIOD                        351


 APPENDICES

 A. PRIMITIVE FOLK-MOOTS                                   381
 B. WATLING STREET AND THE DANELAGH                        382
 C. THE MILITARY ORIGIN OF THE BOROUGHS                    382
 D. THE WORDS "CASTRUM" AND "CASTELLUM"                    383
 E. THE BURGHAL HIDAGE                                     385
 F. THELWALL                                               385
 G. THE WORD "BRETASCHE"                                   386
 H. THE WORD "HURDICIUM"                                   387
 I. THE WORD "HERICIO"                                     388
 K. THE CASTLE OF YALE                                     388
 L. THE CASTLE OF TULLOW                                   389
 M. THE CASTLE OF SLANE                                    390
 N. THE WORD "DONJON"                                      390
 O. THE ARRANGEMENTS IN EARLY KEEPS                        391
 P. KEEPS AS RESIDENCES                                    392
 Q. CASTLES BUILT BY HENRY I.                              392
 R. THE SO-CALLED SHELL KEEP                               393
 S. PROFESSOR LLOYD'S "HISTORY OF WALES"                   393


 SCHEDULE OF ENGLISH CASTLES FROM THE ELEVENTH CENTURY     396


 INDEX                                                     401




LIST OF ILLUSTRATIONS AND PLANS


 FIG.

     Motte-Castles from the Bayeux Tapestry:--Dol, Rennes, Dinan,
     Bayeux, Hastings                                     _Frontispiece_

                                                             FACING PAGE

  1. Typical Motte-Castles:--Topcliffe, Yorks; Laughton-en-le-Morthen,
     Yorks; Anstey, Herts; Dingestow, Monmouth; Hedingham, Essex       4

  2. Anglo-Saxon MS. of Prudentius                                    19

  3. Wallingford, Berks; Wareham, Dorset                              28

  4. Eddisbury, Cheshire; Witham, Essex                               36

  5. Plan of Towcester about 1830                                     42

  6. Shoebury, Essex                                                  52

  7. Willington, Beds                                                 59

  8. Arundel, Sussex; Abergavenny, Monmouth                           98

  9. Barnstaple, Devon; Berkhampstead, Herts; Bishop's Stortford,
     Herts                                                           102

 10. Bourn, Lincs; Bramber, Sussex                                   108

 11. Caerleon, Monmouth; Carisbrooke                                 114

 12. Carlisle; Castle Acre, Norfolk                                  124

 13. Clifford, Hereford; Clitheroe, Lancs; Corfe, Dorset             128

 14. Dover (from a plan in the British Museum, 1756)                 138

 15. Dunster, Somerset; Dudley, Staffs                               144

 16. Durham                                                          146

 17. Ely, Cambs; Ewias Harold, Hereford; Eye, Suffolk                150

 18. Hastings, Sussex; Huntingdon                                    158

 19. Launceston, Cornwall; Lewes, Sussex                             164

 20. Lincoln                                                         166

 21. Monmouth; Montacute, Somerset; Morpeth, Northumberland          168

 22. Norham; Nottingham                                              172

 23. Norwich (from Harrod's _Gleanings among the Castles and
     Convents of Norfolk_, p. 133)                                   174

 24. Okehampton, Devon; Penwortham, Lancs; Pevensey, Sussex          178

 25. Oxford (from _Oxonia Illustrata_, David Loggan, 1675)           180

 26. Pontefract, Yorks; Preston Capes, Northants; Quatford, Salop    188

 27. Rayleigh, Essex; Richard's Castle, Hereford                     192

 28. Richmond, Yorks; Rochester, Kent                                194

 29. Rockingham, Northants                                           202

 30. Old Sarum, Wilts                                                204

 31. Shrewsbury; Skipsea, Yorks                                      208

 32. Stafford; Tamworth, Staffs; Stanton Holgate, Salop; Tickhill,
     Yorks                                                           212

 33. Tonbridge, Kent; Totnes, Devon                                  220

 34. Trematon, Cornwall; Tutbury, Staffs                             226

 35. Wallingford, Berks                                              228

 36. Warwick; Wigmore, Hereford                                      232

 37. Winchester (from a plan by W. Godson, 1750)                     234

 38. Windsor Castle (from Ashmole's _Order of the Garter_)           236

 39. York Castle and Baile Hill (from a plan by P. Chassereau,
     1750)                                                           244

 40. Motte-Castles of North Wales:--Mold, Welshpool, Wrexham,
     Mathraval                                                       260

 41. Motte-Castles of South Wales:--Cilgerran, Blaenporth,
     Chastell Gwalter                                                282

 42. Motte-Castles of South Wales:--Builth, Gemaron, Payn's Castle   290

 43. Motte-Castles of South Wales:--Cardiff, Loughor                 294

 44. Scottish Motte-Castles:--Annan, Moffat, Duffus, Old Hermitage   310

 45. Irish Motte-Castles:--Ardmayle, Downpatrick, Drogheda,
     Castleknock                                                     336




THE EARLY NORMAN CASTLES OF THE BRITISH ISLES




CHAPTER I

INTRODUCTORY


The study of earthworks has been one of the most neglected subjects
in English archæology until quite recent years. It may even be said
that during the first half of the 19th century, less attention was
paid to earthworks than by our older topographical writers. Leland,
in the reign of Henry VIII., never failed to notice the "Dikes and
Hilles, which were Campes of Men of Warre," nor the "Hilles of Yerth
cast up like the Dungeon of sum olde Castelle," which he saw in his
pilgrimages through England. And many of our 17th- and 18th-century
topographers have left us invaluable notices of earthworks which were
extant in their time. But if we turn over the archæological journals of
some fifty years ago, we shall be struck by the paucity of papers on
earthworks, and especially by the complete ignoring, in most cases, of
those connected with castles.

The misfortune attending this neglect, was that it left the ground
open to individual fancy, and each observer formed his own theory of
the earthworks which he happened to have seen, and as often as not,
stated that theory as a fact. We need not be surprised to find Camden
doing this, as he wrote before the dawn of scientific observation;
but that such methods should have been carried on until late in the
19th century is little to the credit of English archæology. Mr Clark's
work on _Mediæval Military Architecture_ (published in 1884), which
has the merit of being one of the first to pay due attention to castle
earthworks, counterbalances that merit by enunciating as a fact a mere
guess of his own, which, as we shall afterwards show, was absolutely
devoid of solid foundation.

The scientific study of English earthworks may be said to have been
begun by General Pitt-Rivers in the last quarter of the 19th century;
but we must not forget that he described himself as a pupil of Canon
Greenwell, whose careful investigations of British barrows form such
an important chapter of prehistoric archæology. General Pitt-Rivers
applied the lessons he had thus learned to the excavation of camps and
dykes, and his labours opened a new era in that branch of research. By
accumulating an immense body of observations, and by recording those
observations with a minuteness intended to forestall future questions,
he built up a storehouse of facts which will furnish materials to all
future workers in prehistoric antiquities. He was too cautious ever to
dogmatise, and if he arrived at conclusions, he was careful to state
them merely as suggestions. But his work destroyed many favourite
antiquarian delusions, even some which had been cherished by very
learned writers, such as Dr Guest's theory of the "Belgic ditches" of
Wiltshire.

A further important step in the study of earthworks was taken by the
late Mr I. Chalkley Gould, when he founded the Committee for Ancient
Earthworks, and drew up the classification of earthworks which is now
being generally adopted by archæological writers. This classification
may be abridged into (_a_) promontory or cliff forts, (_b_) hill forts,
(_c_) rectangular forts, (_d_) moated hillocks, (_e_) moated hillocks
with courts attached, (_f_) banks and ditches surrounding homesteads,
(_g_) manorial works, (_h_) fortified villages.

We venture to think that still further divisions are needed, to include
(1) boundary earthworks; (2) sepulchral or religious circles or
squares; (3) enclosures clearly non-military, intended to protect sheep
and cattle from wolves, or to aid in the capture of wild animals.[2]

This classification, it will be observed, makes no attempt to decide
the dates of the different types of earthworks enumerated. But a great
step forward was taken when these different types were separated
from one another. There had been no greater source of confusion in
the writings of our older antiquaries, than the unscientific idea
that one earthwork was as good as another; that is to say, that one
type of earthwork would do as well as another for any date or any
circumstances. When it is recognised that large classes of earthworks
show similar features, it becomes probable that even if they were not
thrown up in the same historic period, they were at any rate raised to
meet similar sets of circumstances. We may be quite sure that a camp
which contains an area of 60 or 80 acres was not constructed for the
same purpose as one which only contains an area of three.

We are not concerned here, however, with the attempt to disentangle
the dates of the various classes of prehistoric earthworks.[3] Such
generalisations are for the most part premature; and although some
advance is being made in this direction, it is still impossible to
decide without excavation whether a camp of class (_a_) or (_b_)
belongs to the Stone Age, the Bronze Age, or the Iron Age. Our business
is with classes (_d_) and (_e_) of Mr Gould's list, that is, with the
moated hillocks. We shall only treat of the other classes to the extent
which is necessary to bring out the special character of classes (_d_)
and (_e_).

Let us look more closely into these earthworks in their perfect form,
the class (_e_) of the Earthwork Committee's list. They consist, when
fully preserved, of an artificial hillock, 20, 30, 40, or in some rare
instances 100 feet high. The hillock carried a breastwork of earth
round the top, which in many cases is still preserved; this breastwork
enclosed a small court, sometimes only 30 feet in diameter, in rare
cases as large as half an acre; it must have been crowned by a stockade
of timber, and the representations in the Bayeux Tapestry would lead
us to think that it always enclosed a wooden tower.[4] As a rule the
hillock is round, but it is not unfrequently oval, and occasionally
square. The base of the hillock is surrounded by a ditch. Below the
hillock is a court, much larger than the small space enclosed on the
top of the mount. It also has been surrounded by a ditch, which joins
the ditch of the mount, and thus encloses the whole fortification.
The court is defended by earthen banks, both on the scarp and
counterscarp of the ditch, and these banks of course had also their
timber stockades, the remains of which have sometimes been found on
excavation.[5]

[Illustration: FIG. 1.--TYPICAL MOTTE-CASTLES.

Topcliffe, Yorks. Laughton-en-le-Morthen, Yorks. Anstey, Herts.
Dingestow, Monmouth. Hedingham, Essex.]

These are the main features of the earthworks in question. Some
variations may be noticed. The ditch is not invariably carried all
round the hillock, occasionally it is not continued between the hillock
and the court.[6] Sometimes the length of the ditch separating the
hillock from the court is at a higher level than the main ditch.[7]
Often the ditches were evidently dry from the first, but not
infrequently they are wet, and sometimes vestiges of the arrangements
for feeding them are still apparent. The hillock is not invariably
artificial; often it is a natural hill scarped into a conical shape;
sometimes an isolated rock is made use of to serve as a citadel, which
saved much spade-work. The shape of the court is very variable: it may
be square or oblong, with greatly rounded corners, or it may be oval,
or semilunar, or triangular; a very common form is the bean-shaped.
The area covered by these fortifications is much more uniform; one of
the features contrasting them most strongly with the great prehistoric
"camps" of southern England is their comparatively small size. We know
of only one (Skipsea) in which the bailey covers as much as eight
acres; in by far the greater number the whole area included in the
hillock, court, and ditches does not exceed three acres, and often it
is not more than one and a half.[8]

Now this type of fort will tell us a good deal about itself if we
examine it carefully. In the first place, its character is more
pronounced than that of any other class of earthwork. It differs
entirely from the great camps which belong to the tribal period. It
was evidently not designed to accommodate a mass of people with their
flocks and herds. It is small in area, and its citadel, as a rule, is
very small indeed. Dr Sophus Müller, the eminent Danish archæologist,
when dealing with the specimens of this class of fortification which
are to be found in Denmark, made the luminous remark that "the
fortresses of prehistoric times are the defences of the _community_,
north of the Alps as in the old classical lands. Small castles for
an individual and his warrior-band belong to the Middle Ages."[9]
These words give the true direction to which we must turn for the
interpretation of these earthworks.

In the second place, this type presents a peculiar development of
plan, such as we do not expect to find in the earliest times in these
islands. It has a citadel of a most pronounced type. This alone
differentiates it from the prehistoric or Keltic camps which are so
abundant in Great Britain. It might be too hasty a generalisation
to say that no prehistoric camps have citadels, but as a rule the
traverses by which some of these camps are divided appear to have
been made for the purpose of separating the cattle from the people,
rather than as ultimate retreats in time of war. The early German
camps, according to Köhler, have inner enclosures which he thinks
were intended for the residence of the chief; but he calls attention
to the great difference between these camps and the class we are now
considering, in that the inner enclosure is of much greater size.[10]
It would appear that some of the fortifications in England which are
known or suspected to be Saxon have also these inner enclosures of
considerable size (6 acres in the case of Witham), but without any
vestige of the hillock which is the principal feature of class (_e_).

It is clear, in the third place, that the man who threw up earthworks
of this latter class was not only suspicious of his neighbours, but
was even suspicious of his own garrison. For the hillock in the great
majority of cases is so constructed as to be capable of complete
isolation, and capable of defending itself, if necessary, against its
own court. Thus it is probable that the force which followed this
chieftain was not composed of men of his own blood, in whom he could
repose absolute trust; and the earthworks themselves suggest that they
are the work of an invader who came to settle in these islands, who
employed mercenaries instead of tribesmen, and who had to maintain his
settlement by force.

When on further inquiry we find that earthworks of this type are
exceedingly common in France, and are generally found in connection
with feudal castles,[11] and when we consider the area of their
distribution in the United Kingdom, and see that they are to be found
in every county in England, as well as in Wales and in the Normanised
parts of Ireland and Scotland, we see that the Norman invader is the
one to whom they seem to point. We see also that small forts of this
kind, easily and cheaply constructed, and defensible by a small number
of men, exactly correspond to the needs of the Norman invader, both
during the period of the Conquest and for a long time after his first
settlement here.

But it will at once occur to an objector that there have been other
invaders of Britain before the Normans, and it may be asked why these
earthworks were not equally suited to the needs of the Saxon or the
Danish conquerors, and why they may not with equal reason be attributed
to them. To answer this question we will try to discover what kind of
fortifications actually were constructed by the Saxons and Danes, and
to this inquiry we will address ourselves in the succeeding chapters.

It will clear the ground greatly if it is recognised at the outset
that these earthworks are _castles_, in the usual sense of the word;
that is, the private fortified residences of great landowners. It
was the chief merit of Mr G. T. Clark's work on _Mediæval Military
Architecture_, that he showed the perfect correspondence in plan of
these earthen and timber structures with the stone castles which
immediately succeeded them, so that it was only necessary to add a
stone tower and stone walls to these works to convert them into a
Norman castle of the popularly accepted type. We regard the military
character of these works as so fully established that we have not
thought it necessary to discuss the theory that they were temples,
which was suggested by some of our older writers, nor even the more
modern idea that they were moot-hills, which has been defended with
considerable learning by Mr G. L. Gomme.[12] Dr Christison remarks
in his valuable work on Scottish fortifications that an overweening
importance has been attached to moot-hills, without historical
evidence.[13] And Mr George Neilson, in his essay on "The Motes in
Norman Scotland"[14] (to which we shall often have occasion to refer
hereafter), shows that moot-hill in Scotland means nothing but
mote-hill, the hill of the mote or _motte_; but that _moots_ or courts
were held there, just because it had formerly been the site of a
castle, and consequently a seat of jurisdiction.[15]

That some of these hillocks have anciently been sepulchral, we do
not attempt to deny. The Norman seems to have been free from any
superstitious fear which might have hindered him from utilising the
sepulchres of the dead for his personal defence; or else he was
unaware that they were burial-places. There are some very few recorded
instances of prehistoric burials found under the hillocks of castles;
but in ordinary cases, these hillocks would not be large enough for the
_mottes_ of castles.[16] There are, however, some sepulchral barrows
of such great size that it is difficult to distinguish them from
mottes; the absence of a court attached is not sufficient evidence,
as there are some mottes which stand alone, without any accompanying
court. Excavation or documentary evidence can alone decide in these
cases, though the presence of an earthen breastwork on top of the
mount furnishes a strong presumption of a military origin. But the
undoubtedly sepulchral barrows of New Grange and Dowth in Ireland show
signs of having been utilised as castles, having remains of breastworks
on their summits.[17]




CHAPTER II

ANGLO-SAXON FORTIFICATIONS


We have pointed out in the preceding chapter that when it is asked
whether the earthworks of the moated mound-and-court type were the work
of the Anglo-Saxons, the question resolves itself into another, namely,
Did the Anglo-Saxons build castles?

As far as we know, they did not; and although to prove a negative we
can only bring negative evidence, that evidence appears to us to be
very conclusive. But before we deal with it, we will try to find out
what sort of fortifications the Anglo-Saxons actually did construct.

The first fortification which we read of in the _Anglo-Saxon Chronicle_
is that of Bamborough, in Northumberland. The _Anglo-Saxon Chronicle_
tells us that in 547 Ida began to reign in Northumberland, and adds
that he built "Bebbanburh," which was first enclosed with a hedge,
and afterwards with a wall. Unfortunately this celebrated passage is
merely the interpolation of a 12th-century scribe, and is consequently
of no authority whatever,[18] though there is nothing improbable in
the statement, and it is supported by Nennius.[19] Ida's grandson
Ethelfrith gave this fortress to his wife Bebba, from whom it received
the name of Bebbanburh, now Bamborough. It was built without doubt on
the same lofty insulated rock where the castle now stands; for when it
was attacked by Penda in 633, he found the situation so strong that
it was impossible to storm it, and it was only by heaping up wood on
the most accessible side that he was able to set fire to the wooden
stockade.[20] Modern historians talk of this fort as a castle, but all
the older authorities call it a town;[21] nor is there any mention
of a castle at Bamborough till the reign of William II. The area of
the basaltic headland of Bamborough covers 4-3/4 acres, a site large
enough for a city of Ida's day. The church of St Peter was placed on
the highest point. The castle which was built there in Norman times
does not seem to have occupied at first more than a portion of this
site,[22] though it is probable that eventually the townsmen were
expelled from the rock, and that thus the modern town of Bamborough
arose in the levels below. Although 4-3/4 acres may seem a small size
for an _urbs_, it was certainly regarded as such, and was large enough
to protect a considerable body of invaders.

Strange to say, this is the only record which we have of any
fortress-building by the invading Saxons. Until we come to the time
of Alfred, there is hardly an allusion to any fortification in use in
Saxon times.[23] It is mentioned in 571 that the Saxons took four
towns (_tunas_) of the Britons, and the apparent allusion to sieges
seems to show that these British towns had some kind of fortification.
The three _chesters_, which were taken by the Saxons in 577,
Gloucester, Cirencester, and Bath, prove that some Roman cities still
kept their defences. In 755 the slaughter of Cynewulf, king of the West
Saxons, by the etheling Cyneard, is told with unusual detail by the
_Chronicle_. The king was slain in a _bur_ (bower, or isolated women's
chamber[24]), the door of which he attempted to defend; but this _bur_
was itself enclosed in a _burh_, the gates of which were locked by the
etheling who had killed the king, and were defended until they were
forced by the king's avengers. Here it seems to be doubtful whether
the _burh_ was a town or a private enclosure resembling a stable-yard
of modern times. The description of the storming of York by the Danes
in 867 shows that the Roman walls of that city were still preserved.
These passages are the solitary instances of fortifications in England
mentioned by the _Chronicle_ before the time of Alfred.[25] The
invasions of the Danes led at last to a great fortifying epoch, which
preserved our country from being totally overwhelmed by those northern
immigrants.

The little Saxon kingdom of Wessex was the germ of the British Empire.
When Alfred came to the throne it had already absorbed the neighbouring
kingdoms of Kent, Sussex, and Surrey, and the issue hanging in the
balance was whether this small English state would survive the
desolating flood of pagan barbarism which had already overwhelmed the
sister kingdoms of the Midlands and the North. It was given to Alfred
to raise again the fallen standard of Christendom and civilisation, and
to establish an English kingdom on so sound a basis that when, in later
centuries, it successively became the prey of the Dane and the Norman,
the English polity survived both conquests. The wisdom, energy, and
steadfastness of King Alfred and his children and grandchildren were
amongst the most important of the many factors which have helped to
build up the great empire of Britain.

We are concerned here with only one of the measures by which Alfred and
his family secured the triumph of Wessex in her mortal struggle with
the Danes, the fortifications which they raised for the protection of
their subjects. From the pages of the _Anglo-Saxon Chronicle_ we might
be led to think that Alfred's son and daughter, Edward and Ethelfleda,
were the chief builders of fortifications. But there is ample evidence
that they only carried out a systematic purpose which had been
initiated by Alfred. We know that Alfred was a great builder. "What
shall I say," cries Asser, "of the cities and towns which he restored,
and of others which he built which had never existed before! Of the
royal halls and chambers, wonderfully built of stone and wood by his
command!"[26] The _Anglo-Saxon Chronicle_ notices the restoration of
London (886),[27] about which two extant charters are more precise.[28]
It also mentions the building of a work (geweorc) at Athelney, and
another at Limene-muthan (doubtless a repair of the Roman fort at
Lympne), and two works built by Alfred on the banks of the river
Lea.[29] William of Malmesbury tells us that in his boyhood there was
a stone in the nunnery of Shaftesbury which had been taken out of
the walls of the town, which bore this inscription: "Anno dominicæ
incarnationis Alfredus rex fecit hanc urbem, DCCCLXXX, regni sui
VIII."[30] Ethelred, Alfred's son-in-law, built the _burh_ at Worcester
in Alfred's lifetime, as a most interesting charter tells us.[31]

It may be safely assumed, then, that when Edward came to the throne
he found Wessex well provided with defensive places, and that when he
and his sister signalised their conquests in the Midlands by building
strongholds at every fresh step of their advance, they were only
carrying out the policy of their father.

At the time of Alfred's death, and the succession of Edward the Elder
to the crown (901), Ethelfleda, daughter of Alfred, was the wife of
Ethelred, ealdorman of Mercia, who appears to have been a sort of
under-king of that province.[32] On the death of Ethelred in 912,[33]
Edward took possession of London and Oxford and "of all the lands
which owed obedience thereto"--in other words, of that small portion
of Eastern Mercia which was still in English hands; that is, not only
the present Oxfordshire and Middlesex, but part of Herts, part of
Bedfordshire, all Buckinghamshire, and the southern part of Northants.
The Watling Street, which runs north-west from London to Shrewsbury,
and thence north to Chester and Manchester, formed at that time the
dividing line between the English and Danish rule.[34] It would seem
from the course of the story that after Ethelred's death there was some
arrangement between Ethelfleda and her brother, possibly due to the
surrender of the territory mentioned above, which enabled her to rule
English Mercia in greater independence than her husband had enjoyed. Up
to this date we find Edward disposing of the _fyrd_ of Mercia;[35] this
is not mentioned again in Ethelfleda's lifetime. Nothing is clearer,
both from the _Chronicle_ and from Florence, than that the brother and
sister each "did their own," to use an expressive provincial phrase.
Ethelfleda goes her own way, subduing Western Mercia, while Edward
pushes up through Eastern Mercia and Essex to complete the conquest of
East Anglia. A certain concert may be observed in their movements, but
they did not work in company.

The work of fortification begun in Alfred's reign had been continued by
the restoration of the Roman walls of Chester in 908, by Ethelred and
his wife; and Ethelfleda herself (possibly during the lingering illness
which later chroniclers give to her husband) had built a _burh_ at
Bremesbyrig. During the twelve years which elapsed between Ethelred's
death and that of Edward in 924, the brother and sister built no less
than twenty-seven _burhs_, giving a total of thirty, if we add Chester
and Bremesbyrig, and Worcester, which was built in Alfred's reign. Now
what was the nature of these fortifications, which the _Anglo-Saxon
Chronicle_ uniformly calls _burhs_?

There is really not the slightest difficulty in answering this
question. The word is with us still; it is our word _borough_. It is
true we have altered the meaning somewhat, because a borough means
now an enfranchised town; but we must remember that it got that
meaning because the fortified towns, the only ones which were called
_burhs_ or _burgi_, were the first to be enfranchised, and while the
fortifications have become less and less important, the franchise has
become of supreme importance.

Bede, in the earliest times of our history, equated _burh_ with _urbs_,
a city; Alfred in his _Orosius_ translates _civitas_ by _burh_;[36]
the Anglo-Saxon gospels of the 11th century do the same;[36] and
the confederacy of five Danish towns which existed in Mercia in the
10th century is called in contemporary records _fif burga_, the five
boroughs.[37]

_Burh_ is a noun derived from the word _beorgan_, to protect.
Undoubtedly its primitive meaning was that of a _protective enclosure_.
As in the case of the words _tun_, _yard_, or _garth_, and _worth_
or _ward_, the sense of the word became extended from the protecting
bulwark to the place protected. In this sense of a _fortified
enclosure_, the word was naturally applied by the Anglo-Saxons to the
prehistoric and British "camps" which they found in Britain, such as
Cissbury. Moreover, it is clear that some kind of enclosure must have
existed round every farmstead in Saxon times, if only as a protection
against wolves. The illustrated Saxon manuscripts show that the hall
in which the thane dwelt, the ladies' bower, the chapel and other
buildings dependent on the hall, were enclosed in a stockade, and had
gates which without doubt were closed at night.[38] This enclosure may
have been called a _burh_, and the innumerable place-names in England
ending in _borough_ or _bury_[39] seem to suggest that the _burh_
was often nothing more than a stockade, as in so many of these sites
not a vestige of defensive works remains.[40] We may concede that
the original meaning of an _enclosure_ was never entirely lost, and
that it appears to be preserved in a few passages in the Anglo-Saxon
laws. Thus Edmund speaks of _mine burh_ as an asylum, the violation
of which brings its special punishment; and Ethelred II. ordains that
every compurgation shall take place in _thaes kyninges byrig_; and the
_Rectitudines Singularum Personum_ tells us that one of the duties of
the geneat was to build for his lord, and to hedge his _burh_.[41]
But it is absolutely clear that even in these cases a _burh_ was an
enclosure and not a tump; and it is equally clear from the general use
of the word that its main meaning was a _fortified town_. Athelstan
ordains that there shall be a mint in every _burh_; and his laws show
that already the _burh_ has its _gemot_ or meeting, and its _reeve_ or
mayor.[42] He ordains that all _burhs_ are to be repaired fourteen
days after Rogations, and that no market shall be held outside the
town.[43] In the laws of Edgar's time not only the borough-moot and
the borough-reeve are spoken of, but the _burh-waru_ or burgesses.[44]
_Burh_ is contrasted with wapentake as town with country.[45]

[Illustration: FIG. 2.--ANGLO-SAXON MS. OF PRUDENTIUS.

If we wish to multiply proofs that a _burh_ was the same thing as
a borough, we can turn to the Anglo-Saxon illustrated manuscripts,
and we shall find that they give us many pictures of _burhs_, and
that in all cases they are fortified towns.[46] Finally, Florence of
Worcester, one of the most careful of our early chroniclers, who lived
when Anglo-Saxon was still a living language, and who must have known
what a _burh_ meant, translates it by _urbs_ in nineteen cases out
of twenty-six.[47] His authority alone is sufficient to settle this
question, and we need no longer have any doubt that a _burh_ was the
same thing which in mediæval Latin is called a _burgus_, that is a
fortified town, and that our word _borough_ is lawfully descended from
it.

It would not have been necessary to spend so much time on the history
of the word _burh_ if this unfortunate word had not been made the
subject of one of the strangest delusions which ever was imposed on the
archæological world. We refer of course to the theory of the late Mr G.
T. Clark, who contended in his _Mediæval Military Architecture_[48]
that the moated mound of class (_e_), which we have described in our
first chapter, was what the Anglo-Saxons called a _burh_. In other
words, he maintained that the burhs were Saxon castles. It is one
of the most extraordinary and inexplicable things in the history of
English archæology that a man who was not in any sense an Anglo-Saxon
scholar was allowed to affix an entirely new meaning to a very common
Anglo-Saxon word, and that this meaning was at once accepted without
question by historians who had made Anglo-Saxon history their special
study! The present writer makes no pretensions to be an Anglo-Saxon
scholar, but it is easy to pick out the word _burh_ in the _Chronicle_
and the Anglo-Saxon _Laws_, and to find out how the word is translated
in the Latin chronicles; and this little exercise is sufficient in
itself to prove the futility of Mr Clark's contention.

Sentiment perhaps had something to do with Mr Clark's remarkable
success. There is an almost utter lack of tangible monuments of our
national heroes; and therefore people who justly esteemed the labours
of Alfred and his house were pleased when they were told that the
mounds at Tamworth, Warwick, and elsewhere were the work of Ethelfleda,
and that other mounds were the work of Edward the Elder. It did not
occur to them that they were doing a great wrong to the memory of the
children of Alfred in supposing them capable of building these little
earthen and timber castles for their personal defence and that of their
nobles, and leaving the mass of their people at the mercy of the Danes.
Far other was the thought of Ethelfleda, when she and her husband
built the borough of Worcester. As they expressed it in their memorable
charter, it was not only for the defence of the bishop and the churches
of Worcester, but "TO SHELTER ALL THE FOLK."[49] And we may be sure
that the same idea lay at the founding of all the boroughs which were
built by Alfred and by Edward and Ethelfleda. They were to be places
where the whole countryside could take refuge during a Danish raid. The
_Chronicle_ tells us in 894 how Alfred divided his forces into three
parts, the duty of one part being to defend the boroughs; and from
this time forth we constantly find the men of the boroughs doing good
service against the Danes.[50] It was by defending and thus developing
the boroughs of England that Alfred and his descendants saved England
from the Danes.

Thus far we have seen that all the fortifications which we know to have
been built by the Anglo-Saxons were the fortifications of society and
not of the individual. We have heard nothing whatever of the private
castle as an institution in Saxon times; and although this evidence
is only negative, it appears to us to be entitled to much more weight
than has hitherto been given to it. Some writers seem to think that the
private castle was a modest little thing which was content to blush
unseen. This is wholly to mistake the position of the private castle
in history. Such a castle is not merely a social arrangement, it is a
political institution of the highest importance. Where such castles
exist, we are certain to hear of some of them, sooner or later, in the
pages of history.

We can easily test this by comparing Anglo-Saxon history with Norman
of the same period, after castles had arisen in Normandy. Who among
Saxon nobles was more likely to possess a castle than the powerful
Earl Godwin, and his independent sons? Yet when Godwin left the court
of Edward the Confessor, because he would not obey the king's order
to punish the men of Dover for insulting Count Eustace of Boulogne,
we do not hear that he retired to his castle, or that his sons
fortified their castles against the king; we only hear that they met
together at Beverstone (a place where there was no castle before the
14th century)[51] and "arrayed themselves resolutely."[52] Neither
do we hear of any castle belonging to the powerful Earl Siward of
Northumbria, or Leofric, Earl of Mercia. And when Godwin returned
triumphantly to England in 1052 we do not hear of any castles being
restored to him.

Now let us contrast this piece of English history, as told by the
_Anglo-Saxon Chronicle_, with the Norman history of about the same
period, the history of the rebellion of the Norman nobles against their
young duke, William the Bastard. The first thing the nobles do is to
put their castles into a state of defence. William has to take refuge
in the castle of a faithful vassal, Hubert of Rye, until he can safely
reach his own castle of Falaise. After the victory of Val-ès-Dunes,
William had to reduce the castles which still held out, and then to
order the destruction of all the castles which had been erected against
him.[53]

Or let us contrast the _Anglo-Saxon Chronicle_ of 1051 with that of
1088, when certain Norman barons and bishops in England conspired
against the new king, William Rufus. The first thing told us is that
each of the head conspirators "went to his castle, and manned it and
victualled it." Then Bishop Geoffrey makes Bristol Castle the base of
a series of plundering raids. Bishop Wulfstan, on the other hand, aids
the cause of William by preventing an attempt of the rebels on the
castle of Worcester. Roger Bigod throws himself into Norwich Castle,
and harries the shire; Bishop Odo brings the plunder of Kent into his
castle of Rochester. Finally the king's cause wins the day through the
taking of the castles of Tonbridge, Pevensey, Rochester, and Durham.

If we reflect on the contrast which these narratives afford, it surely
is difficult to avoid the conclusion that if the chronicler never
mentions any Saxon castles it is because there were no Saxon castles
to mention. Had Earl Godwin possessed a stronghold in which he could
fortify himself, he would certainly have used it in 1051. And as the
Norman favourites of Edward the Confessor had already begun to build
castles in England, we can imagine no reason why Godwin did not do
the same, except that such a step was impossible to a man who desired
popularity amongst his countrymen. The Welshmen, we are told (that is
the foreigners, the Normans), had erected a castle in Herefordshire
among the people of Earl Sweyn, and had wrought all possible harm
and disgrace to the king's men thereabout.[54] The language of the
_Chronicle_ shows the unpopularity, to say the least of it, of this
castle-building; and one of the conditions which Godwin, when posing
as popular champion, wished to exact from the king, was that the
_Frenchmen who were in the castle_ should be given up to him.[55] When
Godwin returned from his exile, and the Normans took to flight, the
chronicler tells us that some fled west to Pentecost's castle, some
north to Robert's castle. Thus we learn that there were several castles
in England belonging to the Norman favourites.

It is in connection with these Norman favourites that the word _castel_
appears for the first time in the _Anglo-Saxon Chronicle_. This is a
fact of considerable importance in itself; and when we weigh it in
connection with the expressions of dislike recorded above which become
much more explicit and vehement after the Norman Conquest, we cannot
but feel that Mr Freeman's conclusion, that the thing as well as the
word was new, is highly probable.[56] For the hall of the Anglo-Saxon
ealdorman or thane, even when enclosed in an earthwork or stockade,
was a very different thing from the castle of a Norman noble. A castle
is built by a man who lives among enemies, who distrusts his nearest
neighbours as much as any foe from a distance. The Anglo-Saxon noble
had no reason to distrust his neighbours, or to fortify himself against
them. Later historians, who were familiar with the state of things in
Norman times, tell us frequently of castles in the Saxon period; but
it can generally be proved that they misunderstood their authorities.
The genuine contemporary chroniclers of Saxon times never make the
slightest allusion to a Saxon castle.

The word _castellum_, it is true, appears occasionally in Anglo-Saxon
charters, but when it is used it clearly means a town. Thus Egbert of
Kent says in 765: "Trado terram intra castelli moenia supranominati,
id est Hrofescestri, unum viculum cum duobus jugeribus, etc.," where
_castellum_ is evidently the city of Rochester.[57] Offa calls
Wermund "episcopus castelli quod nominatur Hroffeceastre."[58] These
instances can easily be multiplied. Mr W. H. Stevenson remarks that "in
Old-English glosses, from the 8th century Corpus Glossary downwards,
_castellum_ is glossed by _wic_, that is town."[59] In this sense no
doubt we must interpret Asser's "castellum quod dicitur Werham."[60]
Henry of Huntingdon probably meant a town when he says that Edward the
Elder built at Hertford "castrum non immensum sed pulcherrimum." He
generally translates the _burh_ of the _Chronicle_ by _burgus_, and he
shows that he had a correct idea of Edward's work when he says that
at Buckingham Edward "fecit _vallum_ ex utraque parte aquæ"--where
_vallum_ is a translation of _burh_. The difference between a _burh_
and a castle is very clearly expressed by the _Chronicle_ in 1092,
when it says concerning the restoration of Carlisle on its conquest by
William Rufus, "He repaired the borough (burh) and ordered the castle
to be built."

The following is a table of the thirty boroughs built by Ethelfleda and
Edward, arranged chronologically, which will show that we never find a
_motte_, that is a moated mound, on the site of one of these boroughs
unless a Norman castle-builder has been at work there subsequently. The
weak point in Mr Clark's argument was that when he found a motte on a
site which had once been Saxon, he did not stop to inquire what any
subsequent builders might have done there, but at once assumed that the
motte was Saxon. Of course, if we invariably found a motte at _every_
place where Edward or Ethelfleda are said to have built a _burh_, it
would raise a strong presumption that mottes and burhs were the same
thing. But out of the twenty-five burhs which can be identified, in
only ten is there a motte on the same site; and in every case where a
motte is found, except at Bakewell and Towcester, there is recorded
proof of the existence of a Norman castle. In this list, the _burhs_
on both sides of the river at Hertford, Buckingham, and Nottingham
are counted as two, because the very precise indications given in
the _Anglo-Saxon Chronicle_ show that each _burh_ was a separate
construction.


 _Burhs of Ethelfleda._

 Worcester             873-899  A motte and a Norman castle.
 Chester                   908  A motte and a Norman castle.
 Bremesburh                911  Unidentified.
 Scærgate                  913  Unidentified.
 Bridgenorth               913  No motte, but a Norman stone keep.
 Tamworth                  914  A motte and a Norman castle.
 Stafford, N. of Sowe      914  No motte and no Norman castle.
 Eddisbury                 915  No motte and no Norman castle.
 Warwick                   915  A motte and a Norman castle.
 Cyricbyrig (Monk's Kirby) 916  No motte and no Norman castle.
 Weardbyrig                916  Unidentified.
 Runcorn                   916  No motte; a mediæval castle (?).


 _Burhs of Edward the Elder._

 Hertford, N. of Lea       913  No motte and no Norman castle.
 Hertford, S. of Lea       913  A motte and a Norman castle.
 Witham                    914  No motte and no Norman castle.
 Buckingham, S. of Ouse    915  No motte and no Norman castle.
 Buckingham, N. of Ouse    915  A motte and a Norman castle.
 Bedford, S. of Ouse       916  No motte and no Norman castle.
 Maldon                    917  No motte and no Norman castle.
 Towcester                 918  A motte.
 Wigingamere               918  Unidentified.
 Huntingdon                918  A motte and a Norman castle.
 Colchester                918  No motte; an early Norman keep.
 Cledemuthan               918  Unidentified.
 Stamford, S. of Welland   919  No motte and no Norman castle.
 Nottingham, N. of Trent   919  A motte and a Norman castle.
 Thelwall                  920  No motte and no Norman castle.
 Manchester                920  No castle on the ancient site.
 Nottingham, S. of Trent   921  No motte and no Norman castle.
 Bakewell (near to)        921  A motte and bailey.

Out of this list of the _burhs_ of Ethelfleda and Edward, thirteen
are mentioned as boroughs in Domesday Book;[61] and as we ought to
subtract five from the list as unidentified, and also to reckon as
one the boroughs built on two sides of the river, the whole number
should be reduced to twenty-two. So that more than half the boroughs
built by the children of Alfred continued to maintain their existence
during the succeeding centuries, and in fact until the present day.
But the others, for some reason or other, did not take root. Professor
Maitland remarked that many of the boroughs of Edward's day became
rotten boroughs before they were ripe;[62] and it is a proof of the
difficulty of the task which the royal brethren undertook that, with
the exception of Chester, none of the boroughs which they built in the
north-western districts survived till Domesday. In all their boroughs,
except Bakewell, the purpose of defending the great Roman roads and the
main waterways is very apparent.

Our list is very far from being a complete list of all the Anglo-Saxon
boroughs existing in Edward's day. In the document known as the
"Burghal Hidage" we have another quite different list of thirty-two
boroughs,[63] which, according to Professor Maitland, "sets forth
certain arrangements made early in the 10th century for the defence
of Wessex against the Danish inroads."[64] Five at least on the list
are Roman chesters; twenty are mentioned as boroughs in Domesday
Book. There are two among them which are of special interest, because
there is reason to believe that the earthen ramparts which still
surround them are of Saxon origin: Wallingford and Wareham. Both these
fortifications are after the Roman pattern, the earthen banks forming a
square with rounded corners.[65] See Fig. 3.

To complete our knowledge of Anglo-Saxon fortification, we ought
to examine the places mentioned in Anglo-Saxon charters as royal
seats, where possibly defensive works of some kind may have existed.
Unfortunately we are unable to learn that there are any such works,
except at one place, Bensington in Oxfordshire, where about a hundred
years ago "a bank and trench, which seem to have been of a square
form," were to be seen.[66]

[Illustration:FIG. 3. Wallingford, Berks. Wareham, Dorset.]

In the following chapter we shall deal in detail with such archæological
remains as still exist of the boroughs of Edward and Ethelfleda, but
here we will briefly summarise by anticipation the results to which
that chapter will lead. We see that sites defensible by nature were
often seized upon for fortification, as at Bamborough, Bridgenorth, and
Eddisbury; but that this was by no means always the case, as a weak
site, such as Witham, for example, was sometimes rendered defensible by
works which appear to have fulfilled their purpose. In only one case
(Witham) do we find an inner enclosure; and as it is of large size
(9-1/2 acres) it is more probable that the outer enclosure was for
cattle, than that the inner one was designed solely for the protection
of the king and his court. We are not told of stone walls more than
once (at Towcester); but the use of the word _timbrian_, which does
not exclusively mean to build in wood,[67] does not preclude walls of
stone in important places. In the square or oblong form, with rounded
corners, we see the influence which Roman models exercised on eyes
which still beheld them existing.

We see that the main idea of the borough was the same as that of the
prehistoric or British "camp of refuge," in that it was intended for
the defence of society and not of the individual. It was intended to be
a place of refuge for the whole countryside. But it was also something
much more than this, something which belongs to a much more advanced
state of society than the hill-fort.[68] It was a town, a place where
people were expected to live permanently and do their daily work. It
provided a fostering seat for trade and manufactures, two of the chief
factors in the history of civilisation. The men who kept watch and
ward on the ramparts, or who sallied forth in their bands to fight the
Danes, were the men who were slowly building up the prosperity of the
stricken land of England. By studding the great highways of England
with fortified towns, Alfred and his children were not only saving the
kernel of the British Empire, they were laying the sure foundations of
its future progress in the arts and habits of civilised life.




CHAPTER III

ANGLO-SAXON FORTIFICATIONS--_continued_


The bare list which we have given of the boroughs of Edward and
Ethelfleda calls for some explanatory remarks. Let us take first the
boroughs of Ethelfleda.

WORCESTER.--We have already noticed the charter of Ethelred and
Ethelfleda which tells of the building of the burh at Worcester.[69]
There appears to have been a small Roman settlement at Worcester, but
there is no evidence that it was a fortified place.[70] This case lends
some support to the conjecture of Dr Christison, that the Saxons gave
the name of _chester_ to towns which they had themselves fortified.[71]
The mediæval walls of Worcester were probably more extensive than
Ethelfleda's borough, of which no trace remains.

CHESTER is spoken of by the _Anglo-Saxon Chronicle_ in 894 as "a
waste _chester_ in Wirral." It had undoubtedly been a Roman city, and
therefore the work of Ethelred and Ethelfleda here was solely one of
restoration. Brompton, who wrote at the close of the 13th century "a
poor compilation of little authority,"[72] was the first writer to
state that the walls of Chester were enlarged by Ethelfleda so as
to take in the castle, which he fancied to be Roman;[73] and this
statement, being repeated by Leland, has acquired considerable vogue.
It is very unlikely that any extension of the walls was made by the
Mercian pair, seeing that the city was deserted at the time when it
was occupied by the Danes, only fourteen years before. But it is quite
certain that the Norman castle of Chester lay outside the city walls,
as the manor of Gloverstone, which was not within the jurisdiction of
the city, lay between the city and the castle.[74] A charter of Henry
VII. shows that the civic boundary did not extend to the present south
wall in his reign. Ethelfleda's borough probably followed the lines of
the old Roman castrum.

BREMESBYRIG.--This place has not yet been identified. Bromborough on
the Mersey has been suggested, and is not impossible, for the loss of
the _s_ sometimes occurs in place-names; thus Melbury, in Wilts, was
Melsburie in Domesday. Bremesbyrig was the first place restored after
Chester, and as the estuary of the Dee had been secured by the repair
of Chester, so an advance on Bromborough would have for its aim to
secure the estuary of the Mersey. It was outside the Danish frontier
of Watling Street, and could thus be fortified without breach of the
peace in 911. There is a large moated work at Bromborough, enclosing an
area of 10 acres, in the midst of which stands the courthouse of the
manor of Bromborough. But this manor was given by the Earl of Chester
to the monks of St Werburgh about 1152, and it is possible that the
monks fortified it, as they did their manor of Irby in Wirral, against
the incursions of the Welsh. One of the conditions of the Earl's
grant was that the manor is to be maintained in a state of security
and convenience for the holding of the courts appertaining to Chester
Abbey.[75] Thus the fortification appears to be of manorial use, though
this does not preclude the possibility of an earlier origin. On the
other hand, if Bromborough is the same as Brunanburh, where Athelstan's
great battle was fought (and there is much in favour of this), it
cannot possibly have been Bremesbyrig in the days of Edward. Another
site has been suggested by the Rev. C. S. Taylor, in a paper on _The
Danes in Gloucestershire_, Bromsberrow in S. Gloucestershire, one of
the last spurs of the Malvern Hills. Here the top of a small hill has
been encircled with a ditch; but the ditch is so narrow that it does
not suggest a defensive work, and it is remote from any Roman road or
navigable river.

SCERGEAT has not yet been identified. Mr Kerslake argued with some
probability that Shrewsbury is the place;[76] but the etymological
considerations are adverse, and it is more likely that such an
important place as Shrewsbury was fortified before Edward's time.
Leland calls it Scorgate, and says it is "about Severn side."[77] It
should probably be sought within the frontier of Watling Street, which
Ethelfleda does not appear to have yet crossed in 911.

BRIDGENORTH is undoubtedly the Bricge of the _Anglo-Saxon Chronicle_,
as Florence of Worcester identifies it with the Bridgenorth which
Robert Belesme fortified against Henry I. in 1101.[78] Bridgenorth
is on a natural fortification of steep rock, which would only require
a stout wall to make it secure against all the military resources of
the 10th century. We may therefore be quite certain that it was here
Ethelfleda planted her borough, and not (as Mr Eyton unfortunately
conjectured) on the mound outside the city, in the parish of
Oldbury.[79] This mound was far more probably the site of the siege
castle (no doubt of wood) which was erected by Henry I. when he
besieged the city.[80]

TAMWORTH was an ancient city of the Mercian kings, and therefore may
have been fortified before its walls were rebuilt by Ethelfleda.[81]
The line of the ancient town-wall can still be traced in parts, though
it is rapidly disappearing. Dugdale says the town ditch was 45 feet
broad. Tamworth was a borough at the time of Domesday.

STAFFORD has a motte on which stood a Norman castle; but this is not
mentioned in the table, because it stands a mile and a half from the
town on the _southern_ side of the river Sowe, while we are expressly
told by Florence that Ethelfleda's borough was on the _northern_ side,
as the town is now. Stafford was a Domesday borough; some parts of
the mediæval walls still remain. The walls are mentioned in Domesday
Book.[82]

EDDISBURY, in Cheshire (Fig. 4), is the only case in which the work
of Ethelfleda is preserved in a practically unaltered form, as no
town or village has ever grown out of it. The _burh_ stands at the
top of a hill, commanding the junction of two great Roman roads, the
Watling Street from Chester to Manchester, and the branch which it
sends forth to Kinderton on the east. As a very misleading plan of this
work has been published in the _Journal of the British Archæological
Association_ for 1906, the _burh_ has been specially surveyed for this
book by Mr D. H. Montgomerie, who has also furnished the following
description:--

"This plan is approximately oval, and is governed by the shape of
the ground; the work lies at the end of a spur, running S.E. and
terminating in abrupt slopes to the E. and S. The defences on the N.
and W. consist of a ditch and a high outer bank, the proportions of
these varying according to the slope of the hill. There are slight
remains of a light inner rampart along the western half of this side.
The remains of an original entrance (shown in Ormerod's _Cheshire_)
are visible in the middle of the N.W. side, beyond which the ditch and
outer bank have been partially levelled by the encroachments of the
farm buildings. The defences of the S. side seem to have consisted of
a long natural slope, crowned by a steeper scarp, cut back into the
rock, and having traces of a bank along its crest. The S.E. end of the
spur presents several interesting details, for it has been occupied
in mediæval times by a small fortified enclosure, whose defences are
apt to be confused with those of the older Saxon town. The rock makes
a triangular projection at this end, containing the foundations of
mediæval buildings,[83] and strengthened on the N.E. by a slight
ditch some 7 to 10 feet below the crest; the rock on the inner side
of this ditch has been cut back to a nearly vertical face, while on
the outer bank are the footings of a masonry wall extending almost to
the point of the spur. There are traces of another wall defending the
crest on the N.E. and S.; but the base of the triangle, facing the old
enclosure, does not appear to have been strengthened by a cross ditch
or bank.

"It may be noted that this enclosure presents not the slightest
appearance of a motte. It is at a lower level than the body of the
hill, and belongs most certainly to the Edwardian period of the masonry
buildings."

[Illustration: FIG. 4. Eddisbury, Cheshire. Witham, Essex.]

WARWICK Castle has a motte which has been confidently attributed to
Ethelfleda, only because Dugdale copied the assertion of Thomas Rous, a
very imaginative writer of the 15th century, that she was its builder.
The borough which Ethelfleda fortified probably occupied a smaller area
than the mediæval walls built in Edward I.'s reign; and it is probable
that it did not include the site of the castle, as Domesday states
that only four houses were destroyed when the castle was built.[84]
The borough was doubtless erected to protect the Roman road from Bath
to Lincoln, the Foss Way, which passes near it. Domesday Book, after
mentioning that the king's barons have 112 houses in the borough, and
the abbot of Coventry 36, goes on to say that these houses belong to
the lands which the barons hold outside the city, and are rated
there.[85] This is one of the passages from which the late Professor
Maitland concluded that the boroughs planted by Ethelfleda and Edward
were organised on a system of military defence, whereby the magnates in
the country were bound to keep houses in the towns.[86]

CYRICBYRIG.--About this place we adopt the conjecture of Dugdale, who
identified it with Monk's Kirby in Warwickshire, not far from the
borders of Leicestershire, and therefore on the edge of Ethelfleda's
dominions. It lies close to the Foss Way, and about three miles from
Watling Street; like Eddisbury, it is near the junction of two Roman
roads. There are remains of banks and ditches below the church.
Dugdale says "there are certain apparent tokens that the Romans had
some station here; for by digging the ground near the church, there
have been discovered foundations of old walls and Roman bricks."[87]
Possibly Ethelfleda restored a Roman castrum here. At any rate, it
seems a much more likely site than Chirbury in Shropshire, which is
commonly proposed, but which does not lie on any Roman road, and is not
on Ethelfleda's line of advance; nor are there any earthworks there.

WEARDBYRIG has not been identified. Wednesbury was stated by Camden
to be the place,[88] and but for the impossibility of the etymology,
the situation would suit well enough. Weardbyrig must have been an
important place, for it had a mint.[89] Warburton, on the Mersey, has
been gravely suggested, but is impossible, as it takes its name from St
Werburgh.

RUNCORN has not a vestige to show of Ethelfleda's borough; but local
historians have preserved some rather vague accounts of a promontory
fort which once existed at the point where the London and North-Western
Railway bridge enters the river. A rocky headland formerly projected
here into the Mersey, narrowing its course to 400 yards at high water;
a ditch with a circular curve cut off this headland from the shore.
This ditch, from 12 to 16 feet wide, with an inner bank 6 or 7 feet
high, could still be traced in the early part of the 19th century.
Eighteen feet of the headland were cut off when the Duke of Bridgewater
made his canal in 1773, and the ditch was obliterated when the railway
bridge was built. From the measurements which have been preserved, the
area of this fort must have been very small, not exceeding 3 acres at
the outside;[90] and it is unlikely that it represented Ethelfleda's
borough, as the church, which was of pre-Conquest foundation, stood
outside its bounds, and we should certainly have expected to find it
within. As the Norman earls of Chester established a ferry at Runcorn
in the 12th century, and as a castle at Runcorn is spoken of in a
mediæval document,[91] it seems not impossible that there may have
been a Norman castle on this site, as we constantly find such small
fortifications placed to defend a ferry or ford. It is probable that
Ethelfleda's borough was destroyed at an early period by the Northmen,
for Runcorn was not a borough at Domesday, but was then a mere
dependency of the Honour of Halton.


_The Burhs of Edward the Elder._

HERTFORD.--Two burhs were built by Edward at Hertford in 913, one on
the north and the other on the south side of the river Lea. Therefore
if a burh were the same thing as a motte, there ought to be two mottes
at Hertford, one on each side of the river; whereas there is only one,
and that forms part of the works of the Norman castle. Mr Clark, with
his usual confidence, says that the northern mound has "long been laid
low";[92] but there is not the slightest proof that it ever existed
except in his imagination. Hertford was a borough at the time of
Domesday. No earthworks remain.

WITHAM (Fig. 4).--There are some remains of a _burh_ here which are
very remarkable, as they show an inner enclosure within the outer one.
They have been carefully surveyed by Mr F. C. J. Spurrell, who has
published a plan of them.[93] Each enclosure formed roughly a square
with much-rounded corners. The ditch round the outer work was 30 feet
wide; the inner work was not ditched. The area enclosed by the outer
bank was 26-1/4 acres, an enclosure much too large for a castle; the
area of the inner enclosure was 9-1/2 acres. As far as is at present
known, Witham is the only instance we have of an Anglo-Saxon earthwork
which has a double enclosure.[94] Witham is not mentioned as a borough
in Domesday Book, but the fact that it had a mint in the days of
Hardicanute shows that it maintained its borough rights for more than
a hundred years. The name Chipping Hill points to a market within the
borough.

BUCKINGHAM is another case where a _burh_ was built on both sides of
the river, and as at Hertford, there was only one motte, site of the
castle of the Norman Giffards, now almost obliterated. The river Ouse
here makes a long narrow loop to the south-west, within which stands
the town, and, without doubt, this would be the site of Edward's
borough. No trace is left of the second borough on the other side of
the river. Buckingham is one of the boroughs of Domesday.

BEDFORD has had a motte and a Norman castle on the north side of
the Ouse; but this was not the site of Edward's borough, which the
_Chronicle_ tells us was placed on the south side of that river. On the
south side an ancient ditch, 10 or 12 feet broad, with some traces of
an inner rampart, semicircular in plan, but with a square extension,
is still visible, and fills with water at flood times.[95] This is
very likely to be the ditch of Edward's borough. Both at Bedford and
Buckingham the _Chronicle_ states that Edward spent four weeks in
building the _burh_. Mediæval numbers must never be taken as precise;
but the disproportion between four weeks and eight days, the space
often given for the building of an early Norman castle, corresponds
very well to the difference between the time needed to throw up the
bank and stockade of a town, and that needed for the building of an
earthen and wooden castle.

MALDON.--Only one angle of the earthen bank of Edward's borough remains
now, but Gough states that it was an oblong camp enclosing about 22
acres.[96] It had rounded corners and a very wide ditch, with a bank on
both scarp and counterscarp. Maldon was a borough at Domesday;[97] the
king had a hall there, but there was never any castle, nor is there any
trace of a motte.

TOWCESTER (Fig. 5).--There is a motte at Towcester, but no direct
evidence has yet been found for the existence of a Norman castle there,
though Leland says that he was told of "certen Ruines or Diches of a
Castelle."[98] There was a mill and an oven to which the citizens owed
soke,[99] and the value of the manor, which belonged to the king, had
risen very greatly since the Conquest;[100] all facts which render
the existence of a Norman castle extremely likely. But there can be
no question as to the nature of Edward's work at Towcester, as the
_Chronicle_ tells us expressly that "he wrought the burgh at Towcester
with a stone wall."[101] Towcester lies on Watling Street, and is
believed to have been the Roman station of Lactodorum. Baker gives a
plan of the remains existing in his time, which may either be those of
the Roman castrum or of Edward's borough.[102] The area is stated to be
about 35 acres.

WIGINGAMERE.--This place is not yet identified, for the identification
with Wigmore in Herefordshire, though accepted by many respectable
writers, will not stand a moment's examination. Wigmore was entirely
out of Edward's beat, and he had far too much on his hands in 918 to
attempt a campaign in Herefordshire. As Wigingamere appears to have
specially drawn upon itself the wrath of East Anglian and Essex Danes,
it must have lain somewhere in their neighbourhood. The _mere_ which
is included in the name would seem to point to that great inland water
which anciently stretched southwards from the Wash into Cambridgeshire.
The only approach to East Anglia from the south lay along a strip of
open chalk land which lay between the great swamp and the dense forests
which grew east of it.[103] Here ran the ancient road called the
Icknield way. On a peninsula which now runs out into the great fens of
the Cam and the Ouse there is still a village called Wicken, 6 miles
west of the Roman road; and possibly, when the land surrounding this
peninsula was under water, this bight may have been called Wigingamere.
This suggestion of course is merely tentative, but what gives it some
probability is that the Danish army which attacked "the borough at
Wigingamere" came from East Anglia as well as Mercia.[104]

[Illustration: FIG. 5.--PLAN OF TOWCESTER ABOUT 1830.]

HUNTINGDON.--The borough of Huntingdon was probably first built by the
Danes, as it was only repaired by Edward. In Leland's time there were
still some remains of the walls "in places." Huntingdon is one of the
_burgi_ of Domesday.

COLCHESTER.--This of course was a Roman site, and Edward needed only
to restore the walls, as the _Chronicle_ indicates. Colchester
was placed so as to defend the river Colne, just as Maldon defended
the estuary of the Blackwater. As the repair of Colchester and the
successful defence of Wigingamere were followed the same year by the
submission of East Anglia, it seems not unlikely that Edward's various
forces may have made a simultaneous advance, along the coast, and
along the Roman road by the Fen country; but this of course is the
merest conjecture, as the _Chronicle_ gives us no details of this very
important event.

CLEDEMUTHAN.--This place is only mentioned in the Abingdon MS. of the
_Chronicle_, but the year 921 is the date given for its building.
This date should probably be transposed to 918, the year in which,
according to Florence, Edward subjugated East Anglia. It is well known
how confused the chronology of the various versions of the _Anglo-Saxon
Chronicle_ is during the reign of Edward the Elder.[105] Cley, in
Norfolk, would be etymologically deducible from Clede (the _d_ being
frequently dropped, especially in Scandinavian districts), and the
_muthan_ points to some river estuary. Cley is one of the few havens
on the north coast of Norfolk, and its importance in former times was
much greater than now, as is shown not only by the spaciousness of its
Early English church, but by the fact that the port has jurisdiction
for 30 miles along the coast.[106] It would be highly probable that
Edward completed the subjugation of East Anglia by planting a borough
at some important point. But as the real date of the fortification of
Cledemuthan is uncertain, we must be content to leave this matter in
abeyance.[107]

STAMFORD is another case where the _borough_ is clearly said to have
been on the side which is opposite to the one where the Norman castle
stands. Edward's borough was on the south side, the motte and other
remains of the Norman castle are on the north of the Welland. It is
remarkable that the part of Stamford on the south side of the Welland
is still a distinct liberty; it is mentioned in Domesday as the sixth
ward of the borough. The line of the earthworks can still be traced in
parts. The borough on the north side of the Welland was probably first
walled in by the Danes, as it was one of the Five Boroughs--Stamford,
Leicester, Lincoln, Nottingham, and Derby--which appear to have formed
an independent or semi-independent state in middle England.[108]
Stamford is a borough in Domesday.

NOTTINGHAM.--The first mention of a fortress in connection with
Nottingham seems to suggest that it owed its origin to the Danes. In
868 the Danish host which had taken possession of York in the previous
year "went into Mercia to Nottingham, and there took up their winter
quarters. And Burgræd king of Mercia and his Witan begged of Ethelred,
king of the West Saxons, and of Alfred his brother, that they would
help them, that they might fight against the army. And then they went
with the West Saxon force into Mercia as far as Nottingham, and there
encountered the army which was in the fortress (geweorc), and besieged
them there; but there was no great battle fought, and the Mercians made
peace with the army."[109] Nottingham became another of the Danish
Five Boroughs. The Danish host on this occasion came from York, no
doubt in ships down the Ouse and up the Trent. The site would exactly
suit them, as it occupied a very strong position on St Mary's Hill, a
height equal to that on which the castle stands, defended on the south
front by precipitous cliffs, below which ran the river Leen, and only a
very short distance from the junction of the Leen with the Trent, the
great waterway of middle England.[110] Portions of the ancient ditch
were uncovered in 1890, and its outline appears to have been roughly
rectangular, like the Danish camp at Shoebury. The ditch was about 20
feet wide. The area enclosed was about 39 acres.

This borough was captured by Edward the Elder in 919, when after the
death of his sister Ethelfleda he advanced into Danish Mercia, taking
up the work which she had left unfinished.[111] The _Chronicle_ tells
us that he repaired the borough (burh), and garrisoned it with both
English and Danes. Two years later, he evidently felt the necessity of
fortifying the Trent itself, for he built another borough on the south
side of the river, and connected the two boroughs by a bridge, which
must have included a causeway or a wooden stage across the marshes of
the Leen. It is not surprising that the frequent floods of the Trent
have carried away all trace of this second borough.[112] The important
position of Nottingham was maintained in subsequent times, and it was
still a borough at Domesday.

THELWALL.--According to Camden, Thelwall explains by its name the kind
of work which was set up here, a wall composed of the trunks of trees.
This was another attempt to defend the course of the Mersey, which was
once tidal as far as Thelwall. No remains of any fortifications can
now be seen at Thelwall, which was not one of the boroughs which took
root. But the Mersey has changed its course very much at this point,
even before the making of the Ship Canal effected a more complete
alteration.[113]

MANCHESTER.--The _burh_ repaired by Edward the Elder was no doubt
the Roman castrum, which was built on the triangle of land between
the Irwell and the Medlock. Large portions of the walls were still
remaining in Stukeley's time, about 1700, and some fragments have
recently been unearthed by the Manchester Classical Association. It
was one of the smaller kind of Roman stations, its area being only 5
acres. Manchester is not mentioned as a borough in Domesday, but the
old Saxon town was long known as Aldportton, which literally means "the
town of the old city." This is its title in mediæval deeds, and it is
still preserved in _Alport_ Street, a street near the remains of the
_castrum_.[114] The later borough of Manchester, which existed at least
as early as the 13th century, appears to have grown up round the Norman
castle, about a mile from the Roman castrum.[115]

BAKEWELL.--The vagueness of the indication in the _Chronicle_, "nigh
to Bakewell," leaves us in some doubt where we are to look for this
_burh_, which Florence calls an _urbs_. Just outside the village of
Bakewell there are the remains of a motte and bailey castle (a small
motte and bailey of 2 acres), which are always assumed to be the
_burh_ of Edward. But the enclosure is far too small for a borough,
and Edward's burh would certainly have enclosed the church; for though
the present church contains no Saxon architecture, the ancient cross
in the graveyard shows that it stands on a Saxon site. It is more
reasonable to suppose that Edward's borough, if it was at Bakewell,
has disappeared as completely as those of Runcorn, Buckingham, and
Thelwall, and that the motte and bailey belong to one of the many
Norman castles whose names never appear in history. There is no
conclusive evidence for the existence of a Norman castle at Bakewell,
but the names Castle Field, Warden Field, and Court Yard are at least
suggestive.[116] Bakewell was the seat of jurisdiction for the High
Peak Hundred in mediæval times.[117]




CHAPTER IV

DANISH FORTIFICATIONS


We must now inquire into the nature of the fortifications built by the
Danes in England, which are frequently mentioned in the _Anglo-Saxon
Chronicle_. It has often been asserted, and with great confidence, that
the Danes were the authors of the moated mounds of class(_e_); those
in Ireland are invariably spoken of by Lewis in his _Topographical
Dictionary_ as "Danish Raths." This fancy seems to have gone somewhat
out of fashion since Mr Clark's _burh_ theory occupied the field,
though Mr Clark's view is often so loosely expressed as to lead one to
think that he supposed all the Northern nations to be makers of mottes;
in fact, he frequently includes the Anglo-Saxons under the general
title of "Northmen"![118] We must therefore endeavour to find out what
the Danish fortifications actually were.

The _Anglo-Saxon Chronicle_ mentions twenty-four places where the
Danes either threw up fortifications (between 787 and 924) or took up
quarters either for the winter, or for such a period of time that we
may infer that there was some fortification to protect them. The word
used for the fortification is generally _geweorc_, a work, or _fæsten_
(in two places only), which has also the general vague meaning of a
_fastness_. There are ten places where these works or fastnesses are
mentioned in the _Chronicle_:--

1. NOTTINGHAM.--We have already seen that the Danish host took up their
winter quarters here in 868, and that there is the highest probability
that the borough which Edward the Elder restored was first built by
them. We have also seen that it was a camp of roughly rectangular form,
and enclosed a very large area, necessary for great numbers.[119]

2. ROCHESTER.--This city was besieged by the Danes in 885, and they
fortified a camp outside. As the artificial mound called Boley Hill is
outside the city, most topographers have jumped to the conclusion that
this was the Danish camp. But the character of the Danish fortification
is clearly indicated in the _Chronicle_: "they made a work around
themselves," that is, it was an enclosure.[120] They could hardly have
escaped by ship, as they did, if their camp had been above the bridge,
which is known to have existed in Saxon times. But Boley Hill is above
the bridge.

3. MILTON, in Kent (Middeltune).--Hæsten the Dane landed at the mouth
of the Thames with 80 ships, and wrought a _geweorc_ here in 893. Two
places in the neighbourhood of Milton have been suggested as the site
of it, a square earthwork at Bayford Court, near Sittingbourne, and
a very small square enclosure called Castle Rough. Neither of these
are large enough to have been of any use to a force which came in 80
ships.[121] Steenstrup has calculated that the average number of men in
a Viking ship must have been from 40 to 50; Hæsten therefore must have
had at least 3200 men with him. It is therefore probable that the camp
at Milton has been swept away.

4. APPLEDORE.--A still larger Danish force, which had been harrying
the Carlovingian empire, came in 250 ships, with their horses, in 893,
and towed their ships "up the river" (which is now extinct) from Lymne
to Appledore, where they wrought a work. There are no earthworks at
Appledore now, but at Kenardington, 2 miles off, there are remains of
"a roughly defined rectangular work, situated on the north and east of
the church, on the slope of the hill towards the marsh, a very likely
place for an entrenchment thrown up to defend a fleet of light-draught
ships hauled up on the beach."[122] The enclosure was very large, one
side which remains being 600 feet long.[123]

5. BENFLEET.--Here Hæsten wrought a work in 894; here he was defeated
by Alfred's forces, and some of his ships burnt. Mr Spurrell states
that there are still some irregular elevations by the stream and about
the church, which he believes to be remains of the Danish camp.[124]
"As the fleet of ships lay in the Beamfleet, it is obvious that the
camp must have partaken of the character of a fortified _hithe_, with
the wall landward and the shore open to the river and the ships." He
also learned on the spot that when the railway bridge across the Fleet
was being made, the remains of several ancient ships, charred by fire,
and surrounded by numerous human skeletons, were found in the mud.[125]
Benfleet must have been a very large camp, as not only was the joint
army of Danes housed in it, that from Milton and that from Appledore,
but they had with them their wives and children and cattle.

6. SHOEBURY (Fig. 6).--After the storming of the camp at Benfleet by
the Saxon forces, the joint armies of the Danes built another _geweorc_
at Shoebury in Essex. We should therefore expect a large camp here, and
Mr Spurrell has shown that the area was formerly about a third of a
square mile. About half the camp had been washed away by the sea when
Mr Spurrell surveyed it in 1879, but enough was left to give a good
idea of the whole. It was a roughly square rampart, with a ditch about
40 feet wide, the ditch having a kind of berm on the inner side. The
bank also had a slight platform inside, about 3 feet above the general
level.[126] As Hæsten had lost his ships at Benfleet, there would be
no fortified hithe connected with it, and if there had been, the sea
would have swept it away. The camp was abandoned almost as soon as it
was made, and the Danish army started on that remarkable march across
England which the _Saxon Chronicle_ relates. They were overtaken and
besieged by Alfred's forces, in a _fastness_ at

7. BUTTINGTON, on the Severn.--It has sometimes been contended that
this was the Buttington near Chepstow; but as the line of march of
the army was "along the Thames till they reached the Severn, then up
along the Severn,"[127] it is more probable that it was Buttington in
Montgomery, west of Shrewsbury.[128] Here there are remains of a strong
bank with a broad deep ditch, which was evidently part of a rectangular
earthwork, as it runs at right angles to Offa's Dyke, which forms one
side of it. It now encloses both the churchyard and vicarage. Whether
the Danes constructed this earthwork, or found it there, we are not
told.

8. There appear to be no remains of the _geweorc_ on the river Lea,
20 miles above London, made by the Danes in 896. But 20 miles above
London, on the Lea, would land us at Amwell, near Ware. In Brayley's
_Hertfordshire_ it is stated that at Amwell, "on the hill above the
church are traces of a very extensive fortification, the rampart of
which is very distinguishable on the side overlooking the vale through
which the river Lea flows."[129]

[Illustration: FIG. 6. Shoebury, Essex.]

9. BRIDGENORTH, or Quatbridge.--The Winchester MS. of the _Chronicle_
says the Danes wrought a _geweorc_ at Quatbridge, in 896, and passed
the winter there. There is no such place as Quatbridge now, only
Quatford; and seeing there were so few bridges in those days, we are
disposed to accept the statement of the Worcester MS., which must have
been the best informed about events in the west, that Bridgenorth
was the site of their work, especially as the high rock at Bridgenorth
offers a natural fortification. The only circumstance that is in favour
of Quatford is that it is mentioned as a _burgus_ in Domesday, which
shows that it possessed fortifications of the civic kind; and we shall
see later on, that such fortifications were often the work of the
Danes. But this burgus may more probably have been the work of Roger de
Montgomeri, who planted a castle there in the 11th century.

10. TEMPSFORD.--Here the Danes wrought a work in 918.[130] There is a
small oblong enclosure at Tempsford, still in fair preservation, called
Gannock Castle, which is generally supposed to be this Danish work. The
ramparts are about 11 or 12 feet above the bottom of the moat, which is
about 20 feet wide. There is a small circular mound, about 5 feet high,
on top of the rampart, which appears to be so placed as to defend the
entrance. This mound is "edged all round by the root of a small bank,
which may have been the base of a stockaded tower."[131] This curious
little enclosure is different altogether from any of the Danish works
just enumerated, and it is difficult to see what purpose it could have
served. The area enclosed is only half an acre, which would certainly
not have accommodated the large army "from Huntingdon and from the East
Angles," which built the advanced post at Tempsford as a base for the
forcible recovery of the districts which they had lost.[132] Such a
small enclosure as this might possibly have been a citadel, but our
knowledge of Danish camps does not tell us of any with citadels, and it
is hardly likely that the democratic constitution of these pirate bands
would have allowed of a citadel for the chief. It is far more probable
that this work belongs to a later time, and that the Danish camp has
been swept away by the river.[133]

11. READING.--There is no "work" mentioned by the _Anglo-Saxon
Chronicle_ at this place, which the Danes made their headquarters
in 871, but we add it to the list because Asser not only mentions
it, but describes the nature of the fortification. It was a _vallum_
drawn between the rivers Thames and Kennet, so as to enclose a
peninsula.[134] It had several entrances, as the Danes "rushed out from
all the gates" on the Anglo-Saxon attack. Such a fort belongs to the
simplest and easiest kind of defence, used at all times by a general
who is in a hurry, and it has therefore no significance in determining
the general type of Danish works.

Besides these eleven places where _works_ are mentioned, there are
thirteen places where the Danes are said to have taken up their winter
quarters, and where we may be certain that they were protected by some
kind of fortifications. These are Thanet, Sheppey, Thetford, York,
London, Torkesey, Repton, Cambridge, Exeter, Chippenham, Cirencester,
Fulham, and Mersey Island. Four places out of this list--York, London,
Exeter, and Cirencester--were Roman _castra_, whose walls were still
available for defence. Three--Thanet, Sheppey, and Mersey--were
islands, and thus naturally defended, being much more insular than they
are now.[135] Three--Thetford, Torkesey, and Cambridge--appear as
burgi in Domesday, showing that they were fortified towns. It is highly
probable that the Danes threw up the first fortifications of these
boroughs. There are no remains of town banks at Torkesey; at Cambridge
the outline of the town bank can be traced in places;[136] and at
Thetford there was formerly an earthwork on the Suffolk side of the
river, which appears to have formed three sides of a square, abutting
on the river, and enclosing the most ancient part of the town.[137]
Chippenham and Repton were ancient seats of the Anglo-Saxon kings, and
may have had fortifications, but nothing remains now. Chippenham is
a borough by prescription, therefore of ancient date. At Fulham, on
the Thames, there is a quadrangular moat and bank round the Bishop of
London's palace, which is sometimes supposed to be the camp made by the
Danes in 879; but it may equally well be mediæval. There was formerly a
harbour at Fulham.[138]

It must be confessed that this list of Danish fortresses furnishes us
with a very slender basis for generalisation as to the nature of Danish
fortifications, judging from the actual remains. All we can say is that
in six cases out of twenty-four (not including Tempsford or Fulham) the
work appears to have been rectangular. In the case of Shoebury, about
which we have the best evidence, the imitation of Roman models seems
to be clear. If we turn from remaining facts to _à priori_ likelihoods,
we call to mind that the Danes were a much-travelled people, had been
in Gaul as well as in England, and had had opportunities of observing
Roman fortifications, as well as much practice both in the assault and
defence of fortified places. It may not be without significance that it
is not until after the return of "the army" from France that we hear of
their building camps at all, except in the case of Reading.

As far as our information goes, their camps were without citadels. What
evidence we have from the other side of the channel supports the same
conclusion. Richer gives us an account of the storming of a fortress
of the Northmen at Eu, by King Raoul, in 925, from which it is clear
that as soon as the king's soldiers had got over the vallum, they were
masters of the place; there was no citadel to attack.[139] Dudo speaks
of the Vikings "fortifying themselves, after the manner of a _castrum_,
by heaped up earth-banks drawn round themselves," and it is clear from
the rest of his description that the camp had no citadel.[140]

In no case do we find anything to justify the theory that mottes were
an accompaniment of Danish camps. In five cases out of the twenty-four
there are or were mottes at the places mentioned, but in all cases
they belonged to Norman castles. The magnificent motte called the
Castle Hill at Thetford was on the opposite side of the river to the
borough, which we have seen reason to think was the site of the Danish
winter quarters. Torkesey in Leland's time had by the river side "a
Hille of Yerth cast up," which he judged to be the donjon of some old
castle, probably rightly, though we have been unable as yet to find any
mention of a Norman castle at Torkesey; a brick castle of much more
recent date is still standing near the river, and probably the motte to
which Leland alludes was destroyed when this was built. The motte at
Cambridge is placed inside the original bounds of the borough, and was
part of the Norman castle.[141] We have already dealt with the Boley
Hill at Rochester, and shall have more to say about it hereafter. The
rock motte at Nottingham was probably not cut off by a ditch from the
rest of the headland until the Norman castle was built.

[Illustration: FIG. 7. Willington, Beds.]

It seems highly probable that besides providing accommodation in their
camps for very large numbers of people, the Danes sometimes fortified
the hithes where they drew up their ships on shore, or even constructed
fortified harbours.[142] We have already quoted Mr Spurrell's remark
on the hithe[143] at Benfleet (p. 51), and there is at least one place
in England which seems to prove the existence of fortified harbours.
This is Willington, on the river Ouse, in Bedfordshire, which has been
carefully described by Mr A. R. Goddard.[144] This "camp" consists
of two wards, and a wide outer enclosure (Fig. 7). "But one of the
most interesting features is the presence of two harbours, contained
within the defences and communicating with the river." Mr Goddard
points out that the dimensions of the smaller one are almost the same
as those of the "nausts" (ship-sheds or small docks) of the Vikings in
Iceland. He also cites from the _Jomsvikinga Saga_ the description of
a harbour made by the Viking Palnatoki at Jomsborg. "There he had a
large and strong sea _burg_ made. He also had a harbour made within the
_burg_ in which 300 long ships could lie at the same time, all being
locked within the burg." The harbours at Willington are large enough
to accommodate between twenty-five and thirty-five ships of the Danish
type. Unfortunately there is no historical proof that the Willington
works were Danish, though their construction makes it very likely. Nor
have any works of a similar character been as yet observed in England,
as far as we are aware.

But if archæology and topography give a somewhat scanty answer to our
question about the nature of Danish fortifications, there are other
fields of research, opened up of late years, from which we can glean
important facts, bearing directly on the subject which we are treating.
Herr Steenstrup's exhaustive inquiry into the Danish settlement in
England has shown that the way in which the Danes maintained their hold
on the northern and eastern shires was by planting fortified towns on
which the soldiers and peasants dwelling around were dependent.[145]
The _Anglo-Saxon Chronicle_ gives us a glimpse of these arrangements
when it speaks of the Danes who owed obedience to Bedford, Derby,
Leicester, Northampton, and Cambridge.[146] It also tells us of the
Five Boroughs, which, as we have already said, appear to have been a
confederation of boroughs forming an independent Danish state between
the Danish kingdoms of East Anglia and Northumbria.

The same system was followed by the Danes who colonised Ireland.
"The colony had a centre in a fortified town, or it consisted almost
exclusively of dwellers in one. But round this town was a district,
in which the Irish inhabitants had to pay taxes to the lords of the
town."[147] The Irish chronicle called _The Wars of the Gaedhil and the
Gaill_ says, further, that Norse soldiers were quartered in the country
round these towns in the houses of the native Irish, and it even says
that there was hardly a house without a Norseman.[148] Herr Steenstrup
does not go so far as to assert that this system of quartering obtained
in England also; but he shows that it is probable, and we may add
that such a system would help to explain the speedy absorption of the
Danes into the Anglo-Saxon population, which took place in the Danelaw
districts.[149]

The large numbers of the Danish forces, and the fact that in the second
period of their invasions they brought their wives and children with
them, would render camps of large area necessary. These numbers alone
make it ridiculous to attribute to the Danes the small motte castles of
class (_e_), whose average area is not more than 3 acres.

Finally, the Danish host was not a feudal host. Steenstrup asserts
that the principle of the composition of the host was the voluntary
association of equally powerful leaders, of whom one was chosen
as head, and was implicitedly obeyed, but had only a temporary
authority.[150] We should not, therefore, expect to find the Danish
camps provided with the citadels by which the feudal baron defended his
personal safety. When Rollo and his host were coming up the Seine, the
Frankish king Raoul sent messengers to ask them who they were, and what
was the name of their chief. "Danes," was the reply, "and we have no
chief, for we are all equal."[151] That such an answer would be given
by men who were following a leader so distinguished as Rollo shows the
spirit of independence which pervaded the Danish hosts, and how little
a separate fortification for the chief would comport with their methods
of warfare.[152]

We may conclude, then, with every appearance of certainty that the
Danish camps were enclosures of large area which very much resembled
the larger Roman _castra_, and that, like these, they frequently
grew into towns. Placed as they generally were on good havens, or
on navigable rivers, they were most suitable places for trade; and
it turned out that the Danes, who were a people of great natural
aptitudes, had a special aptitude for commerce.[153] Dr Cunningham
remarks that they were the leading merchants of the country, and he
attributes to them a large share in the development of town life in
England.[154] The organisation of their armies was purely military, but
at the same time democratic; and when it was applied to a settled life
in the new country, the organisation of the town was the form which it
took. The Lagmen of Lincoln, Stamford, Cambridge, Chester, and York are
a peculiarly Scandinavian institution, which we find still existing at
the time of the Domesday Survey.[155]

Thus we see that the fortifications of the Danes, like those of
the Anglo-Saxons, were the fortifications of the community. And we
shall see in the next chapter that this was the general type of the
fortifications which were being raised in Western Europe in the 9th
century.




CHAPTER V

THE ORIGIN OF PRIVATE CASTLES


We have now seen that history furnishes no instance of the existence
of private castles among the Anglo-Saxons or the Danes (previous
to the arrival of Edward the Confessor's Norman friends), and we
have endeavoured to show that this negative evidence is of great
significance. If, assuming that we are right in accepting it as
conclusive, we ask why the Anglo-Saxons did not build private castles,
the answer is ready to hand in the researches of the late Dr Stubbs,
the late Professor Maitland, Dr J. H. Round, and Professor Vinogradoff,
which have thrown so much fresh light on the constitutional history
of England. These writers have made it clear that whatever tendencies
towards feudalism there were in England before the Conquest, the system
of military tenure, which is the backbone of feudalism, was introduced
into England by William the Conqueror.[156] "Feudalism, in both
tenure and government was, so far as it existed in England, brought
full-grown from France," says Dr Stubbs; and this statement is not
merely supported, but strengthened, by the work of the later writers
named.[157] The institutions of the Anglo-Saxons, when they settled in
England, were tribal; and though these institutions were in a state
of decay in the 11th century, they were not completely superseded by
feudal institutions till after the Norman Conquest.

We should naturally expect, then, that the fortifications erected by
the Anglo-Saxons would be those adapted to their originally tribal
state, that is, in the words which we have so often used already,
they would be those of the community and not of the individual. And
as far as we can discover the character of these fortifications, we
find that this was actually the case. As we have seen, we find one of
the earliest kings, Ida, building for the defence of himself and his
followers what Bede calls a city; and we find Alfred and his children
also building and repairing cities, at the time of the Danish invasions.

The same kind of thing was going on at about the same time in Germany
and in France. Henry the Fowler (919-936), that great restorer of the
Austrasian kingdom, planted on the frontiers which were exposed to
the attacks of the Danes and Huns a number of walled strongholds, not
only for the purpose of resisting invasion, but to afford a place of
refuge to all the inhabitants of the country. He ordained that every
ninth man of the peasants in the district must build for himself and
his nine companions a dwelling in the "Burg," and provide barns and
storehouses, and that the third part of all crops must be delivered
and housed in these towns.[158] In this way, says the historian
Giesebrecht, he sought to accustom the Saxons, who had hitherto dwelt
in isolated farms, or open villages, to life in towns. He ordered that
all assemblies of the people should be held in towns. Giesebrecht also
remarks that it is not improbable that Henry the Fowler had the example
of Edward the Elder of England before his eyes when he established
these rows of frontier towns.[159]

The same causes led, on Neustrian soil, to the fortification of a
number of cities, the walls of which had fallen into decay during the
period of peace before the invasions of the Danes. Thus Charles the
Bald commanded Le Mans and Tours to be fortified "as a defence _for
the people_ against the Northmen."[160] The bishops were particularly
active in thus defending the people of their dioceses. Archbishop Fulk
rebuilt the walls of Rheims, between 884 and 900;[161] his successor,
Hervey, fortified the town of Coucy[162] (about 900); the Bishop of
Cambray built new walls to his city in 887-911;[163] and Bishop Erluin
fortified Peronne in 1001, "as a defence against marauders, and a
refuge for the husbandmen of the country."[164] But permission had
probably to be asked in all these cases, as it certainly had in the
last. The Carlovingian sovereigns represented a well-ordered state,
modelled on the pattern of the Roman Empire; they were jealous of any
attempts at self-defence which did not proceed from the State, and thus
as long as they had the power they strove to put down all associations
or buildings of a military character which did not emanate from their
imperial authority.

The history of the 9th and 10th centuries is the history of the gradual
break-up of the Carlovingian Empire, and the rise of feudalism on
its ruins. In 877, the year of his death, Charles the Bald signed a
decree making the counts of the provinces, who until then had been
imperial officers, hereditary. He thus, as Sismondi says, annihilated
the remains of royal authority in the provinces.[165] The removable
officers now became local sovereigns. Gradually, as the Carlovingian
Empire fell to pieces, the artificial organisation of the feudal system
arose to take its place. By the end of the 10th century the victory of
feudalism was complete; and the victory of feudalism was the victory of
the private castle.

"The very word castle," says Guizot, "brings with it the idea of
feudal society; we see it rising before us. It was feudalism that
built these castles which once covered our soil, and whose ruins are
still scattered upon it. They were the declaration of its triumph.
Nothing like them had existed on Gallo-Roman soil. Before the Germanic
invasion, the great landed proprietors dwelt either in the cities, or
in beautiful houses agreeably situated near the cities."[166] These
Gallo-Roman villas had no fortifications;[167] nor were the Roman
villas in England fortified.[168] It was the business of the State to
defend the community; this was the theory so long sustained by imperial
Rome, and which broke down so completely under the later Carlovingians.

In the time of Charlemagne and Louis le Debonnaire, even the royal
palaces do not appear to have been fortified. They were always spoken
of as _palatia_, never as _castella_. The Danes, when they took
possession of the palace of Nimeguen in 880, fortified it with ditches
and banks.[169] Charles the Bald appears to have been the first to
fortify the palace of Compiègne.[170]

Although there can be no doubt that private castles had become
extremely common on the mainland of Western Europe before the end of
the 10th century, it is more difficult than is generally supposed to
trace their first appearance. Historians, even those of great repute,
have been somewhat careless in translating the words _castrum_ or
_castellum_ as _castle_ or _château_, and taking them in the sense of
the feudal or private castle.[171] We have already pointed out that
these words in our Anglo-Saxon charters mean a town or village.[172]
The fact is that from Roman times until toward the end of the 9th
century the words _castrum_ and _castellum_ are used indifferently
for a fortified city or town, or a temporary camp. The expression
_civitates et castella_ is not uncommon, and might lead one to think
that a distinction was drawn between large and small towns, or forts.
But it is far more likely that it is a mere pleonasm, a bit of that
redundancy which was always dear to the mediæval scribe who was trying
to write well. For as the instances cited in the Appendix will prove,
we constantly find the words _castrum_ and _castellum_ used for the
same town, sometimes even in the same paragraph. Later, from the last
quarter of the 9th century to the middle of the 12th century, these
same words are used indifferently for a town or a castle, and it is
impossible to tell, except by the context, whether a town or a castle
is meant; and often even the context throws no light upon it.

This makes it extremely difficult to say with any exactness when the
private castle first arose. We seem indeed to have a fixed date in
the Capitulary of Pistes, issued by Charles the Bald in 864,[173] in
which he straightly ordered that all who had made castles, forts, or
hedge-works without his permission should forthwith be compelled to
destroy them, because through them the whole neighbourhood suffered
depredation and annoyance. This edict shows, we might argue, that
private castles were sufficiently numerous by the year 864 to have
become a public nuisance, calling for special legislation. But the
chronicles of the second half of the 9th century do not reveal any
extensive prevalence of private castles. Indeed, after studying all
the most important chronicles of Neustria and Austrasia during this
period, the present writer has only been able to find four instances
of fortifications which have any claim at all to be considered private
castles; and even this claim is doubtful.[174]

When we come to the chroniclers of the middle of the 10th century
we find a marked difference. It is true that the words _castrum_,
_castellum_, _municipium_, _oppidum_, _munitio_, are still used quite
indifferently by Flodoard and other writers for one and the same thing,
and that in a great many cases they obviously mean a fortified town.
But there are other cases where they evidently mean a castle. And if
we compare these writers with the earlier ones in the same way as we
have already compared the pre-Conquest portion of the _Anglo-Saxon
Chronicle_ with the chroniclers of the 11th and 12th centuries, we
find the same contrast between them. In the pages of Flodoard or Ademar
the action constantly turns on the building, besieging, and burning of
castles, which by whatever name they are called, have every appearance
of being private castles. In fact before we get to the end of the
century, the private castle is as much the leading feature of the drama
as it is in the 11th or 12th centuries.

Why, then, had the chroniclers no fresh word for a thing which
was in its essential nature so novel? The obvious and only answer
is that the private castle in its earlier stages was nothing
more than an embankment with a wooden stockade thrown round some
_villa_ or farm belonging to a private owner, and was therefore
indistinguishable in appearance, though radically different in idea,
from the fortifications which had hitherto been thrown up for the
protection of the community.[175] How easily we may be mistaken in
the meaning of the word _castellum_, if we interpret it according
to modern ideas, may be seen by comparing the account of the bridge
built by Charlemagne over the Elbe, in the _Annales Laurissenses_,
with Eginhards narrative of the same affair. The former states that
Charlemagne built a _castellum_ of wood and earth at each end of the
bridge, while the latter tells us that it was a _vallum_ to protect
a garrison which he placed there. This, however, was a work of
public utility, and not a private castle. But scanty as the evidence
is, it all leads us to infer that the first private castles were
fortifications of this simple nature.[176] Mazières-on-the-Meuse, which
was besieged for four weeks by Archbishop Hervey, took its name from
the _macerias_ or banks which Count Erlebald had constructed around
it. It is impossible to say whether this enclosure should be called a
castle or a town, but in idea it was certainly a castle, since it was
an enclosure formed for private, not for public interests.

Whether these first private castles were provided with towers we have
no evidence either to prove or to disprove. No instance occurs from
which we can conclude that they possessed any kind of citadel, before
the middle of the 10th century.[177] But before the century is far
advanced, we hear of towers in connection with the great towns, which,
whether they were originally mural towers or not, are evidently private
strongholds, and may justly be called keeps. The earliest instance
known to the writer is in 924, when the tower of the _presidium_
where Herbert Count of Vermandois had imprisoned Charles the Simple
was burnt accidentally.[178] This tower must have been restored, as
nine years later it withstood a six weeks' siege from King Raoul. A
possibly earlier instance is that of Nantes, where Bishop Fulcher had
made a castle in 889; for when this castle was restored by Count Alan
Barbetorte (937-943), we are told that he _restored_ the principal
tower and made it into his own house.[179] Count Herbert built a keep
in Laon before 931; and this appears to have been a different tower to
the one attached to the royal house which Louis d'Outremer had built
at the gate of the city.[180] We hear also of towers at Amiens (950),
Coucy (958), Chalons (963), and Rheims (988). All these towers, it will
be observed, are connected with towns.[181] The first stone keep in the
country for whose date we have positive evidence, is that of Langeais,
built by Fulk Nerra, Count of Anjou, about the year 994; its ruins
still exist.

But we are concerned more particularly here with the origin of
the motte-and-bailey castle. The exact place or time of its first
appearance is still a matter of conjecture. Certainly there is not a
word in the chronicles which is descriptive of this kind of castle
before the beginning of the 11th century.[182] The first historical
mention of a castle which is clearly of the motte-and-bailey kind is
in the Chronicle of St Florent le Vieil, where, at a date which the
modern biographer of Fulk Nerra fixes at 1010, we learn that this
same Count of Anjou built a castle on the western side of the hill
Mont-Glonne, at St Florent le Vieil, on the Loire, and threw up an
_agger_ on which he built a wooden tower.[183] In this case the word
_agger_ evidently means a motte. But Fulk began to reign in 987;
he was a great builder of castles, and was famed for his skill in
military affairs.[184] One of his first castles, built between 991 and
994, was at Montbazon, not far from Tours. About 500 metres from the
later castle of Montbazon is a motte and outworks, which De Salies
not unreasonably supposes to be the original castle of Fulk.[185]
Montrichard, Chateaufort, Chérament, Montboyau, and Baugé are all
castles built by Fulk, and all have or had mottes. Montboyau is the
clearest case of all, as it was demolished by Fulk a few years after he
built it, and has never been restored, so that the immense motte and
outworks which are still to be seen remain very much in their original
state, except that a modern tower has been placed on the motte, which
is now called Bellevue.[186]

It was a tempting theory at one time to the writer to see in Fulk Nerra
the inventor of the motte type of castle, for independently of his fame
in military architecture, he is the first mediæval chieftain who is
known to have employed mercenary troops.[187] Now as we have already
suggested in Chapter I., the plan of the motte-and-bailey castle
strongly suggests that there may be a connection between its adoption
and the use of mercenaries. For the plan of this kind of castle seems
to hint that the owner does not only mistrust his enemies, he also does
not completely trust his garrison. The keep in which he and his family
live is placed on the top of the motte, which is ditched round so as to
separate it from the bailey; the provisions on which all are dependent
are stored in the cellar of the keep, so that they are under his own
hand; and the keys of the outer ward are brought to him every night,
and placed under his pillow.[188]

But unfortunately for this theory, there is some evidence of the
raising of mottes at an earlier period in the 10th century than the
accession of Fulk Nerra. Thibault-le-Tricheur, who was Count of Blois
and Chartres from 932 to 962, was also a great builder, and it is
recorded of him that he built the keeps of Chartres, Chateaudun,[189]
Blois, and Chinon,[190] and the castle of Saumur; these must have been
finished before 962. Now there was anciently a motte at Blois, for in
the 12th century, Fulk V. of Anjou burnt the whole fortress, "_except
the house on the motte_."[191] There was also a motte at Saumur;[192]
and the plan of the castle of Chinon is not inconsistent with the
existence of a former motte.[193] These instances seem to put back the
existence of the motte castle to the middle of the 10th century.

We know of no earlier claim than this, unless we were to accept the
statement of Lambert of Ardres that Sigfrid the Dane, who occupied the
county of Guisnes about the year 928, fortified the town, and enclosed
his own _dunio_ with a double ditch.[194] If this were true, we have a
clear instance of a motte built in the first half of the 10th century.
But Lambert's work was written at the end of the 12th century, with the
object of glorifying the counts of Guisnes, and its editor regards
the early part of it as fabulous. That Sigfrid fortified the _town_ of
Guisnes we can easily believe, as we know the Danes commonly did the
like (see Chapter IV.); but that he built himself a personal castle is
unlikely.[195]

It is the more unlikely, because the Danes in Normandy do not appear
to have built personal castles until the feudal system was introduced
there by Richard Sans Peur. The settlement in Normandy was not on
feudal lines. "Rollo divided out the lands among his powerful comrades,
and there is scarcely any doubt that they received these lands as
inheritable property, without any other pledge than to help Rollo in
the defence of the country."[196] "The Norman constitution at Rollo's
death can be described thus, that the duke ruled the country as an
independent prince in relation to the Franks; but for its internal
government he had a council at his side, whose individual members felt
themselves almost as powerful as the duke himself."[197] Sir Francis
Palgrave asserts that feudalism was introduced into Normandy by the
Duke Richard Sans Peur, the grandson of Rollo, towards the middle of
the 10th century. He "enforced a most extensive conversion of allodial
lands into feudal tenure," and exacted from his baronage the same
feudal submission which he himself had rendered to Hugh Capet.[198]

It is quite in accordance with this that in the narrative of Dudo, who
is our only authority for the history of Normandy in the 10th century,
there is no mention of a private castle anywhere. We are told that
Rollo restored the walls and towers of the _cities_ of Normandy,[199]
and it is clear from the context that the _castra_ of Rouen, Fécamp,
and Evreux, which are mentioned, are fortified cities, not castles.
Even the ducal residence at Rouen is spoken of as a _palatium_ or an
_aula_, not as a castle; and it does not appear to have possessed a
keep until (as we are told by a later writer) the same Duke Richard
who introduced the feudal system into Normandy built one for his own
residence.[200] It is possible that when the feudal oath was exacted
from the more important barons, permission was given to them to build
castles for themselves; thus we hear from Ordericus of the castle of
Aquila, built in the days of Duke Richard; the castle of the lords
of Grantmesnil at Norrei; the castle of Belesme; all of which appear
to have been private castles.[201] But there seems to have been no
general building of castles until the time of William the Conqueror's
minority, when his rebellious subjects raised castles against him
on all sides. "Plura per loca aggeres erexerunt, et tutissimas sibi
munitiones construxerunt."[202] It is generally, and doubtless
correctly, supposed that _aggeres_ in this passage means mottes, and
taking this statement along with the great number of mottes which
are still to be found in Normandy, it has been further assumed (and
the present writer was disposed to share the idea) that this was the
time of the first invention of mottes. But the facts which have
been now adduced, tracing back the first known mottes to the time of
Thibault-le-Tricheur, and the county of Blois, show that the Norman
claim to the invention of this mode of fortification must be given up.
If the Normans were late in adopting feudalism, they were probably
equally late in adopting private castles, and the fortifications of
William I.'s time were most likely copied from castles outside the
Norman frontier.[203]

It might be thought that the general expectation of the end of the
world in the year 1000, which prevailed towards the end of the 10th
century, had something to do with the spread of these wooden castles,
as it might have seemed scarcely worth while to build costly structures
of stone. But it is not necessary to resort to this hypothesis,
because there is quite sufficient evidence to show that long before
this forecast of doom was accepted, wood was a very common, if not
the commonest, material used in fortification. The reader has only to
open his Cæsar to see how familiar wooden towers and wooden palisades
were to the Romans; and he has only to study carefully the chronicles
of the 9th, 10th, 11th, and 12th centuries to see how all-prevalent
this mode of fortification continued to be. The general adoption of
the feudal system must have brought about a demand for cheap castles,
which was excellently met by the motte with its wooden keep and its
stockaded bailey. M. Enlart has pointed out that wooden defences
have one important advantage over stone ones, their greater cohesion,
which enabled them to resist the blows of the battering-ram better
than rubble masonry.[204] Their great disadvantage was their liability
to fire; but this was obviated, as in the time of the Romans, by
spreading wet hides over the outsides. Stone castles were still built,
where money and means were available, as we see from Fulk Nerra's keep
at Langeais; but the devastations of the Northmen had decimated the
population of Gaul; labour must have been dear, and skilled masons hard
to find. In these social and economic reasons we have sufficient cause
for the rapid spread of wooden castles in France.

The sum of the evidence which we have been reviewing is this:
the earliest mottes which we know of were _probably_ built by
Thibault-le-Tricheur about the middle of the 10th century. But in the
present state of our knowledge we must leave the question of the time
and place of their first origin open. The only thing about which we can
be certain is that they were the product of feudalism, and cannot have
arisen till it had taken root; that is to say, not earlier than the
10th century.




CHAPTER VI

DISTRIBUTION AND CHARACTERISTICS OF MOTTE-CASTLES


The motte-and-bailey type of castle is to be found throughout feudal
Europe, but is probably more prevalent in France and the British
Isles than anywhere else. We say _probably_, because there are as
yet no statistics prepared on which to base a comparison.[205] How
recent the inquiry into this subject is may be learned from the fact
that Krieg von Hochfelden, writing in 1859, denied the existence of
mottes in Germany;[206] and even Cohausen in 1898 threw doubt upon
them,[207] although General Köhler in 1887 had already declared that
"the researches of recent years have shown that the motte was spread
over the whole of Germany, and was in use even in the 13th and 14th
centuries."[208] The greater number of the castles described by Piper
in his work on Austrian castles are on the motte-and-bailey plan,
though the motte in those mountainous provinces is generally of natural
rock, isolated either by nature or art. Mottes were not uncommon in
Italy, according to Muratori,[209] and are especially frequent in
Calabria, where we may strongly suspect that they were introduced by
the Norman conqueror, Robert Guiscard.[210] It is not improbable that
the Franks of the first crusade planted in Palestine the type of castle
to which they were accustomed at home, for several of the excellent
plans in Rey's _Architecture des Croisés_ show clearly enough the
motte-and-bailey plan.[211] In most of these cases the motte was a
natural rock.

On the other hand, we are told by Köhler that motte-castles are not
found among the Slavonic nations, because they never adopted the feudal
system.[212] Nor are there any in Norway or Sweden.[213] Denmark
has some, which are attributed by Dr Sophus Müller to the mediæval
period.[214]

Of course whenever a motte was thrown up, the first castle upon it
must have been a wooden one. A stone keep could not be placed on
loose soil.[215] The motte, therefore, must always represent the
oldest castle. But there is no reason to think that the motte and
its wooden keep were merely temporary expedients, intended always to
be replaced as soon as possible by stone buildings. Even after stone
castles had been fully developed, wood continued to hold its ground as
a solid building material until a very late period.[216] And mottes
were used not only throughout the 11th and 12th centuries, but even
as late as the 13th. King John built many castles of this type in
Ireland; and as late as 1242 Henry III. ordered a motte and wooden
castle to be built in the island of Rhé.[217] Muratori gives a much
later instance: in 1320 Can Grande caused a great motte to be built
near Pavia, and surrounded with a ditch and hedge, in order to build
a castle on it.[218] And as will be seen in the next chapter, there
is considerable evidence that many mottes in England which were set up
in the reign of William I., retained their wooden towers or stockades
even till as late as the reign of Edward I. The motte at Drogheda held
out some time against Cromwell, and is spoken of by him as a very
strong place, having a good graft (ditch) and strongly palisaded.[219]
Tickhill Castle in Yorkshire had a palisade on the counterscarp of the
ditch when it was taken by Cromwell.[220]

The position of these motte-castles is wholly different from that
of prehistoric fortresses. They are almost invariably placed in the
arable country, and as a rule not in isolated situations, but in the
immediate neighbourhood of towns or villages. It is rare indeed to
find a motte-castle in a wild, mountainous situation in England. The
only instance which occurs to the writer is that of the motte on the
top of the Hereford Beacon; but there is great probability that this
was a post fortified by the Bishop of Hereford in the 13th century to
protect his game from the Earl of Gloucester. Nothing pointing to a
prehistoric origin was found in this motte when it was excavated by Mr
Hilton Price,[221] though the camp in which it is placed is supposed to
be prehistoric.

The great majority of mottes in England are planted either on or near
Roman or other ancient roads, or on navigable rivers.[222] It was
essential to the Norman settlers that they should be near some road
which would help them to visit their other estates, which William had
been so careful to scatter, and would also enable them to revisit from
time to time their estates in Normandy.[223] The rivers of England were
much fuller of water in mediæval times than they are now, and were much
more extensively used for traffic; they were real waterways. When we
find a motte perched on a river which is not navigable, the purpose
probably was to defend some ford, or to exact tolls from passengers.
Thus the Ferry Hill (corrupted into Fairy Hill) at Whitwood stands at
the spot where the direct road from Pontefract to Leeds would cross the
Calder. It was probably not usual for the motte to be dependent on a
stream or a spring for its supply of water, and this is another point
in which the mediæval castle differs markedly from the prehistoric
camp; wells have been found in a number of mottes which have been
excavated, and it is probable that this was the general plan, though we
have not sufficient statistics on this subject as yet.[224]

Occasionally, but very rarely, we find two mottes in the same castle.
The only instances in England known to the writer are at Lewes and
Lincoln.[225] It is not unfrequent to find a motte very near a stone
castle. In this case it is either the abandoned site of the original
wooden castle, or it is a siege castle raised to blockade the other
one. We constantly hear of these siege castles being built in the
Middle Ages; their purpose was not for actual attack, but to watch the
besieged fort and prevent supplies from being carried in.[226] Hillocks
were also thrown up for the purpose of placing _balistæ_ and other
siege engines upon them; but these would be much smaller than mottes,
and would be placed much nearer the walls than blockade castles.

The mottes of France are in all probability much more decidedly
military than those of England. France was a land of private war, after
the dissolution of the empire of Charlemagne; and no doubt one of the
reasons for the rapid spread of the motte-castle, after its invention,
was due to the facilities which it offered for this terrible game.
In England the reasons for the erection of mottes seem to have been
manorial rather than military; that is, the Norman landholder desired
a safe residence for himself amidst a hostile peasantry, rather than
a strong military position which could hold out against skilful and
well-armed foes.

Attached to the castle, both in England and abroad, we frequently
find an additional enclosure, much larger than the comparatively
small area of the bailey proper. This was the _burgus_ or borough,
which inevitably sprang up round every castle which had a lengthened
existence. Our older antiquaries, finding that the word _burgenses_
was commonly used in Domesday in connection with a site where a
castle existed, formed the mistaken idea that a _burgus_ necessarily
implied a castle. But a _burgus_ was the same thing as a _burh_, that
is, a _borough_ or fortified town. It may have existed long before the
castle, or it may have been created after the castle was built. The
latter case was very common, for the noble who built a castle would
find it to his advantage to build a _burgus_ near it.[227] In exchange
for the protection offered by the borough wall or bank, he could demand
_gablum_ or rent from the burghers; he could compel them to grind their
corn at his mill, and bake their bread at his oven; he could exact
tolls on all commodities entering the borough; and if there was a
market he would receive a certain percentage on all sales. The borough
was therefore an important source of revenue to the baron. Domesday
Book mentions the _new borough_ at Rhuddlan, evidently built as soon
as the castle had been planted on the deserted banks of the Clwydd. In
some cases a "new borough" is clearly a new suburb, doubtless having
its own fortifications, built specially for the protection of the
Norman settlers in England, as at Norwich and Nottingham.[228]

That even in the 12th century a motte was considered an essential
feature of a castle is shown by Neckham's treatise "De Utensilibus,"
where he gives directions as to how a castle should be built; the motte
should be placed on a site well defended by nature; it should have a
stockade of squared logs round the top; the keep on the motte should
be furnished with turrets and battlements, and crates of stones for
missiles should be always provided, as well as a perpetual spring of
water, and secret passages and posterns, by which help might reach the
besieged.[229]

What the outward appearance of these motte-castles was we learn from
the Bayeux Tapestry, which gives us several instructive pictures of
motte-castles existing in the 11th century at Dol, Rennes, Dinan,
and Bayeux.[230] There is considerable variety in these pictures,
and something no doubt must be ascribed to fancy; but all show the
main features of a stockade round the top of the motte, enclosing a
wooden tower, a ditch round the foot of the motte, with a bank on
the counterscarp, and a stepped wooden bridge, up which horses were
evidently trained to climb, leading across the moat to the stockade of
the motte. In no case is the bailey distinctly depicted, but we may
assume that it has been already taken, and that the horsemen are riding
over it to the gate-house which (in the picture of Dinan) stands at the
foot of the bridge. The towers appear to be square, but in the case of
Rennes and Bayeux, are surmounted by a cupola roof. Decoration does not
appear to be have been neglected, and the general appearance of the
buildings, far from being of a makeshift character, must have been very
picturesque.

The picture of the building of the motte at Hastings shows only a
stockade on top of the motte; this may be because the artist intended
to represent the work as incomplete. What is remarkable about this
picture is that the motte appears to be formed in layers of different
materials. We might ascribe this to the fancy of the embroiderer,
were it not that layers of this kind have occasionally been found
in mottes which have been excavated or destroyed. Thus the motte at
Carisbrook, which was opened in 1903, was found to be composed of
alternate layers of large and small chalk rubble. In some cases, layers
of stones have been found; in others (as at York and Burton) a motte
formed of loose material has been cased in a sort of pie-crust of
heavy clay. In the Castle Hill at Hallaton in Leicestershire layers of
peat and hazel branches, as well as of clay and stone boulders, were
found. But our information on this subject is too scanty to justify any
generalisations as to the general construction of mottes.

The pictures shown in the Bayeux Tapestry agree very well with the
description given by a 12th-century writer of the castle of Merchem,
near Dixmüde, in the life of John, Bishop of Terouenne, who died in
1130. "Bishop John used to stay frequently at Merchem when he was
going round his diocese. Near the churchyard was an exceedingly high
fortification, which might be called a castle or _municipium_, built
according to the fashion of that country by the lord of the manor many
years before. For it is the custom of the nobles of that region, who
spend their time for the most part in private war, in order to defend
themselves from their enemies to make a hill of earth, as high as they
can, and encircle it with a ditch as broad and deep as possible. They
surround the upper edge of this hill with a very strong wall of hewn
logs, placing towers on the circuit, according to their means. Inside
this wall they plant their house, or keep (arcem), which overlooks
the whole thing. The entrance to this fortress is only by a bridge,
which rises from the counterscarp of the ditch, supported on double
or even triple columns, till it reaches the upper edge of the motte
(agger)."[231] The chronicler goes on to relate how this wooden bridge
broke down under the crowd of people who were following the bishop, and
all fell 35 feet into the ditch, where the water was up to their knees.
There is no mention of a bailey in this account, but a bailey was so
absolutely necessary to a residential castle, in order to find room
for the stables, lodgings, barns, smithies and other workshops, which
were necessary dependencies of a feudal household, that it can seldom
have been omitted, and the comparatively rare instances which we find
of mottes which appear never to have had baileys were probably outposts
dependent on some more important castle.

Lambert of Ardres, the panegyrist of the counts of Guisnes,[232]
writing about 1194, gives us a minute and most interesting description
of the wooden castle of Ardres, built about the year 1117. "Arnold,
lord of Ardres, built on the motte of Ardres a wooden house, excelling
all the houses of Flanders of that period both in material and in
carpenter's work. The first storey was on the surface of the ground,
where were cellars and granaries, and great boxes, tuns, casks, and
other domestic utensils. In the storey above were the dwelling and
common living rooms of the residents, in which were the larders,
the rooms of the bakers and butlers, and the great chamber in which
the lord and his wife slept. Adjoining this was a private room, the
dormitory of the waiting maids and children. In the inner part of the
great chamber was a certain private room, where at early dawn or in
the evening or during sickness or at time of blood-letting, or for
warming the maids and weaned children, they used to have a fire.... In
the upper storey of the house were garret rooms, in which on the one
side the sons (when they wished it) on the other side the daughters
(because they were obliged) of the lord of the house used to sleep. In
this storey also the watchmen and the servants appointed to keep the
house took their sleep at some time or other. High up on the east side
of the house, in a convenient place, was the chapel, which was made
like unto the tabernacle of Solomon in its ceiling and painting. There
were stairs and passages from storey to storey, from the house into the
kitchen, from room to room, and again from the house into the _loggia_
(logium), where they used to sit in conversation for recreation, and
again from the loggia into the oratory."[233]

This description proves that these wooden castles were no mere rude
sheds for temporary occupation, but that they were carefully built
dwellings designed for permanent residence. The description is useful
for the light it throws on the stone keeps whose ruins remain to us.
They probably had very similar arrangements, and though only their
outside walls are now existing, they must have been divided into
different rooms by wooden partitions which have now perished.[234]

In this account of Lambert's it is further mentioned that the kitchen
was joined to the house or keep, and was a building of two floors, the
lower one being occupied by live stock, while the upper one was the
actual kitchen. We must remember that this account was written at the
end of the 12th century. In the earlier and simpler manners of the 11th
century it is probable that the cooking was more generally carried on
in the open air, as it was among the Anglo-Saxons.[235] The danger of
fire would prevent the development of chimneys in wooden castles; we
have seen that there was only one in this wonderful castle of Ardres.
But even after stone castles became common, we have evidence that the
kitchen was often an isolated building in the courtyard. One such
kitchen still exists in the monastic ruins of Glastonbury.

The word _mota_, which was used in the 12th century for the artificial
hills on which the wooden keeps of these castles were placed, comes
from an old French word _motte_, meaning a clod of earth, which is
still used in France for a small earthen hillock.[236] The keep itself
appears to have been called a _bretasche_, though this word seems to
have meant a wooden tower of any kind, and was used both for mural
towers and for the movable wooden towers employed for sieges.[237] At a
much later period it was given to the wooden balconies by which walls
were defended, but the writer has found no instance of this use of the
word before the 14th century. On the contrary, these wooden galleries
for the purpose of defending the foot of the walls by throwing missiles
down are called _hurdicia_ or hourdes in the documents, a word of
cognate origin to our word _hoarding_.[238] The word _bretasche_ is
also of Teutonic origin, akin to the German _brett_, a board.

The court at the base of the hillock is always called the _ballium_,
_bayle_, or _bailey_, a word for which Skeat suggests the Latin
_baculus_, a stick, as a possible though very doubtful ancestor. The
wooden wall which surrounded this court was the _palum_, _pelum_, or
_palitium_ of the documents, a word which Mr Neilson has proved to be
the origin of the _peels_ so common in Lowland Scotland, though it has
been mistakenly applied to the towers enclosed by these peels.[239]
The _palitium_ was the stockade on the inner bank of the ditch which
enclosed the bailey; but the outer or counterscarp bank had also its
special defence, called the _hericio_, from its bristling nature
(French _hérisson_, a hedgehog). There can be little doubt that it
was sometimes an actual hedge of brambles, at other times of stakes
intertwined with osiers or thorns.[240]

Thus the words most commonly used in connection with these wooden
castles are chiefly French in form, but a French that is tinctured with
Teutonic blood. This is just what we might expect, since the first
castles of feudalism arose on Gallic soil (France or Flanders), but on
soil which was ruled by men of Teutonic descent. We may regard it as
fairly certain that it was in the region anciently known as Neustria
that the motte-castle first appeared; and as we have previously shown,
there is some reason to think that the centre of that region was
the place where it originated. But this must for the present remain
doubtful. What we regard as certain is that it was from France, and
from Normandy in particular, that it was introduced into the British
Isles; and to those islands we must now turn.




CHAPTER VII

THE CASTLES OF THE NORMANS IN ENGLAND


In this chapter we propose to give a list, in alphabetical order for
convenience of reference, of the castles which are known to have
existed in England in the 11th century, because they are mentioned
either in Domesday Book, or in charters of the period, or in some
contemporary chronicle.[241] We do not for a moment suppose that this
catalogue of eighty-four castles is a complete list of those which
were built in England in the reigns of William I. and William II. We
have little doubt that all the castles in the county towns, such as
Leicester, Northampton, and Guildford, and those which we hear of first
as the seats of important nobles in the reign of Henry II., such as
Marlborough, Groby, Bungay, Ongar, were castles built shortly after the
Conquest, nearly all of them being places which have (or had) mottes.
Domesday Book only mentions fifty castles in England and Wales,[242]
but it is well known that the Survey is as capricious in its mention
of castles as in its mention of churches. It is possible that further
research in charters which the writer has been unable to examine may
furnish additional castles, but the list now given may be regarded as
complete as far as materials generally accessible will allow.[243] One
of the castles mentioned (Richard's Castle) and probably two others
(Hereford and Ewias) existed before the Conquest; they were the work of
those Norman friends of Edward the Confessor whom he endowed with lands
in England.

Out of this list of eighty-four castles we shall find that no less than
seventy-one have or had mottes. The exceptions are the Tower of London,
Colchester, Pevensey, and Chepstow, where a stone keep was part of the
original design, and a motte was therefore unnecessary: Bamborough,
Peak, and Tynemouth, where the site was sufficiently defended by
precipices: Carlisle and Richmond, whose original design is unknown to
us: Belvoir, Dover, Exeter, and Monmouth, which might on many grounds
be counted as motte-castles, but as the evidence is not conclusive,
we do not mark them as such; but even if we leave them out, with the
other exceptions, we shall find that nearly 86 per cent. of our list of
castles of the 11th century are of the motte-and-bailey type.

About forty-three of these castles are attached to towns. Of these,
less than a third are placed inside the Roman walls or the Saxon or
Danish earthworks of the towns, while at least two-thirds are wholly or
partly outside these enclosures.[244] This circumstance is important,
because the position outside the town indicates the mistrust of an
invader, not the confidence of a native prince. In the only two cases
where we know anything of the position of the residence of the Saxon
kings we find it in the middle of the city.[245] Even when the castle
is inside the town walls it is almost invariably close to the walls, so
that an escape into the country might always be possible.[246]

Of the towns or manors in which these castles were situated, Domesday
Book gives us the value in King Edward's and King William's time in
sixty-two instances. In forty-five cases the value has risen; in twelve
it has fallen; in five it is stationary. Evidently something has caused
a great increase of prosperity in these cases, and it can hardly be
anything else than the impetus given to trade through the security
afforded by a Norman castle.

Our list shows that Mr Clark's confident statement, that the moated
mounds were the centres of large and important estates in Saxon times,
was a dream. Out of forty-one mottes in country districts, thirty-six
are found in places which were quite insignificant in King Edward's
day, and only five can be said to occupy the centres of important Saxon
manors.[247]

In the table in the Appendix, the area occupied by the original
baileys of the castles in this list has been measured accurately by a
planimeter, from the 25-in. Ordnance maps, in all cases in which that
was possible.[248] This table proves that the early Norman castles
were very small in area, suitable only for the personal defence of a
chieftain who had only a small force at his disposal, and absolutely
unsuited for a people in the tribal state of development, like the
ancient Britons, or for the scheme of national defence inaugurated by
Alfred and Edward. We may remark here that in not a single case is any
masonry which is certainly early Norman to be found on one of these
mottes; where the date can be ascertained, the stonework is invariably
later than the 11th century.

ABERGAVENNY (Fig. 8).--This castle, being in Monmouthshire, must be
included in our list. The earliest notice of it is a document stating
that Hamelin de Ballon gave the church and chapel of the castle of
Abergavenny, and the land for making a _bourg_, and an oven of their
own, to the Abbey of St Vincent at Le Mans.[249]

The castle occupies a pointed spur at the S. end of the town, whose
walls converge so as to include the castle as part of the defence.
The motte has been much altered during recent years, and is crowned
by a modern building; but a plan in _Coxe's Tour in Monmouthshire_,
1800, shows it in its original round form. The bailey is roughly of a
pentagonal shape, covering 1 acre, and is defended by a curtain wall
with mural towers and a gatehouse. The ditch on the W. and N. is much
filled in and obscured by the encroachment of the town. On the E.
the ground descends in a steep scarp, which merges into those of the
headland on which the motte is placed.[250]

[Illustration: FIG. 8. Arundel, Sussex. Abergavenny, Monmouth.]

ARUNDEL (Fig. 8).--"The castrum of Arundel," says Domesday Book, "paid
40_s._ in King Edward's time from a certain mill, and 20_s._ from
three boardlands (or feorm-lands), and 2_s._ from one pasture. Now,
between the town feorm and the water-gate and the ships' dues, it pays
12_l._"[251] _Castrum_ in Domesday nearly always means a castle; yet
the description here given is certainly that of a town and not of a
castle. We must therefore regard it as an instance of the fluctuating
meaning which both _castrum_ and _castellum_ had in the 11th
century.[252] Arundel is one of the towns mentioned in the "Burghal
Hidage."[253] But even accepting that the description in Domesday
refers to the town, we can have very little doubt that the original
earthen castle was reared by Roger de Montgomeri, to whom William
I. gave the Rapes of Arundel and Chichester, and whom he afterwards
made Earl of Shrewsbury.[254] Roger had contributed sixty ships to
William's fleet, and both he and his sons were highly favoured and
trusted by William, until the sons forfeited that confidence. We shall
see afterwards that their names are connected with several important
castles of the early Norman settlement. We shall see also that the
Rapes into which Sussex was divided--Chichester, Arundel, Bramber,
Lewes, Pevensey, and Hastings--were all furnished with Norman castles,
each with the characteristic motte, except Pevensey, which had a stone
keep. Each of these castles, at the time of the Survey, defended a port
by which direct access could be had to Normandy. It was to protect his
base that William fortified these important estuaries, and committed
them to the keeping of some of the most prominent of the Norman leaders.

The castle stands on the end of a high and narrow ridge of the South
Downs, above the town of Arundel. It consists of an oblong ward,
covering 4-1/2 acres, in the middle of which, but on the line of the
west wall, is a large motte, about 70 feet high, surrounded by its own
ditch. The lower and perhaps original bailey is only 2 acres in extent.
Round the top of the motte is a slightly oval wall, of the kind called
by Mr Clark a shell keep. We have elsewhere expressed our doubts of
the correctness of this term.[255] In all the more important castles
we find that the keep on top of the motte has a small ward attached to
it, and Arundel is no exception to this rule; it has the remains of a
tower, as well as the wall round the motte. The tower is a small one,
but it is large enough for the king's chamber in times which were not
extravagant in domestic architecture. It is probable that this tower,
and the stone wall round the motte are the work of Henry II., as he
spent nearly 340_l._ on this castle between the years 1170 and 1187.
His work consisted chiefly of a wall, a king's chamber, a chapel, and
a tower.[256] The wall of the motte corresponds in style to the work
of the middle of his reign; it is built of flints, but cased with Caen
stone brought from Normandy, and has Norman buttresses. The original
Norman doorway on the south side (now walled up) has the chevron
moulding, which shows that it is not earlier than the 12th century.
The tower, which we may assume to be the tower of Henry II.'s records,
has a round arched entrance, and contains a chapel and a chamber (now
ruined) besides a well chamber.

There is earlier Norman work still remaining in the bailey, namely,
the fine gateway, which though of plain and severe Norman, is larger
and loftier than the early work of that style, and of superior
masonry.[257] The one Pipe Roll of Henry I. which we possess shows that
he spent 78_l._ 6_s._ 2_d._ on the castle in 1130, and possibly this
refers to this gatehouse.[258] We know that Henry was a great builder,
but so was the former owner of this castle, Robert Belesme, son of
Roger de Montgomeri.

The value of the town of Arundel had greatly increased since the
Conquest, at the time of the Domesday Survey.[259]

BAMBOROUGH, Northumberland.--We first hear of this castle in the reign
of Rufus, when it was defended against the king by Robert Mowbray, the
rebel Earl of Northumberland; but there can be little doubt that the
earliest castle on this natural bastion was built in the Conqueror's
reign. In the 13th century certain lands were held by the tenure of
supplying wood to the castle of Bamborough, and it was declared that
this obligation had existed ever since the time of William I.[260]
William certainly found no castle there, for Bamborough had fallen
into utter ruin and desolation by the middle of the 11th century.[261]
William's hold on Northumberland was too precarious to give opportunity
for so long and costly a work as the building of a stone keep. It is
more probable that a strong wooden castle was the fortress of the
governors of Northumberland under the first Norman kings, and that
the present stone keep was built in Henry II.'s reign.[262] There is
no motte at Bamborough, nor was one needed on a site which is itself
a natural motte, more precipitous and defensible than any artificial
hill.[263] As the Domesday Survey does not extend to Northumberland, we
have no statement of the value of Bamborough. The area of the castle is
4-3/4 acres.

BARNSTAPLE, Devon (Fig. 9).--This castle is not mentioned in Domesday,
but the town belonged to Judhael, one of the followers of the
Conqueror, whose name suggests a Breton origin. William gave him large
estates in Devon and Cornwall. A charter of Judhael's to the priory
which he founded at Barnstaple makes mention of the castle.[264]
Barnstaple, at the head of the estuary of the Taw, was a borough at
Domesday, and the castle was placed inside the town walls.[265] The
motte remains in good condition; the winding walks which now lead to
the top are certainly no part of the original plan, but are generally
found in cases where the motte has been incorporated in a garden. There
was formerly a stone keep, of which no vestige remains.[266] The castle
seems to have formed the apex of a town of roughly triangular shape.
The bailey can just be traced, and must have covered 1-1/3 acres.

The former value of Barnstaple is not given in the Survey, so we cannot
tell whether it had risen or not.

[Illustration: FIG. 9. Barnstaple, Devon. Bishop's Stortford, Herts.
Berkhampstead, Herts.]

BELVOIR, Leicester.--This castle was founded by the Norman Robert
de Todeni, who died in 1088.[267] It stands on a natural hill, so
steep and isolated that it might be called a natural motte. The first
castle was destroyed by King John, and the modernising of the site has
entirely destroyed any earthworks which may have existed on the hill.
There appears to have been a shell wall, from the descriptions
given by Nicholls and Leland.[268] It was situated in the manor of
Bottesdene, a manor of no great importance, but which had risen in
value at the date of the Survey.[269]

BERKELEY, or NESS.--The identity of Berkeley Castle with the Ness
castle of Domesday may be regarded as certain. All that the Survey says
about it is: "In Ness there are five hides belonging to Berkeley, which
Earl William put out to make a little castle."[270] Earl William is
William FitzOsbern, the trusty friend and counsellor of the Conqueror,
who had made him Earl of Herefordshire. He had also authority over the
north and west of England during William's first absence in Normandy,
and part of the commission he received from William was to build
castles where they were needed.[271] Berkeley was a royal manor with
a large number of berewicks, and the probable meaning of the passage
in Domesday is that Earl William removed the _geldability_ of the
five hides occupying the peninsula or _ness_ which stretches from
Berkeley to the Severn, bounded on the south by the Little Avon, and
appropriated these lands to the upkeep of a small castle. This castle
can hardly have been placed anywhere but at Berkeley, for there is
no trace of any other castle in the district.[272] Earl Godwin had
sometimes resided at Berkeley, but probably his residence there was the
monastery which by evil means had come into his hands;[273] for we
never hear of any castle in connection with Godwin. But a Norman motte
exists at Berkeley, though buried in the stone shell built by Henry II.
Mr Clark remarks: "If the masonry of Berkeley Castle were removed, its
remains would show a mound of earth, and attached to three sides of it
a platform, the whole encircled with a ditch or scarp."[274] The motte
raised by Earl William has, in fact, been revetted with a stone shell
of the 12th century, whose bold chevron ornament over the entrance
gives evidence of its epoch. What is still more remarkable is that
documentary evidence exists to fix the date of this transformation.
A charter of Henry II. is preserved at Berkeley Castle, in which he
grants the manor to Robert Fitzhardinge, pledging himself at the same
time to fortify a castle there, according to Robert's wish.[275]
Robert's wish probably was to possess a stone keep, like those which
had been rising in so many places during the 12th century. But there
had been a Norman lord at Berkeley before Fitzhardinge, Roger de
Berkeley, whose representatives only lost the manor through having
taken sides with Stephen in the civil war.[276] This Roger no doubt
occupied the wooden castle on the motte built by William FitzOsbern.
Henry II.'s shell was probably the first masonry connected with the
castle. This remarkable keep is nearly circular, and has three round
turrets and one oblong. As the latter, Thorpe's Tower, was rebuilt in
Edward III.'s reign, it probably took the place of a round tower. The
keep is built of rubble, and its Norman buttresses (it has several
later ones) project about a foot. The cross loopholes in the walls are
undoubtedly insertions of the time of Edward III. The buildings in the
bailey are chiefly of the time of Edward III., but the bailey walls
have some Norman buttresses, and are probably of the same date as the
keep.[277] This bailey is nearly square, and the motte, which is in one
corner, encroaches upon about a quarter of it. The small size of the
area which it encloses, not much more than half an acre, corresponds to
the statement of Domesday Book that it was "a little castle." There is
no trace of the usual ditch surrounding the motte, and the smallness of
the bailey makes it unlikely that there ever was one. A second bailey
has been added to the first,[278] and the whole is surrounded on three
sides by a moat, the fourth side having formerly had a steep descent
into swamps, which formed sufficient protection.[279]

There is no statement in the Survey of the value of Ness, but the whole
manor of Berkeley had risen since the Conquest.[280]

BERKHAMPSTEAD, Herts (Fig. 9).--Mr D. H. Montgomerie rightly calls
this a magnificent example of an earthwork fortress.[281] It is first
mentioned in a charter of Richard I., which recapitulates the original
charter of William, son of Robert, Count of Mortain, in which he gives
the chapel of this castle to the Abbey of Grestein in Normandy.[282]
We may, therefore, with all probability look upon this as one of the
castles built by the Conqueror's half-brother. And this will account
for the exceptional strength of the work, which comprises a motte 40
feet high, ditched round (formerly), and a bailey of 2-2/3 acres,
surrounded not only with the usual ditch and banks, but with a second
ditch outside the counterscarp bank, which encircles both motte and
bailey. At two important points in its line, this counterscarp bank
is enlarged into mounds which have evidently once carried wooden
towers;[283] if this arrangement belonged to the original plan, as it
most probably did, it confirms a remark which we have made elsewhere as
to the early use of wooden mural towers. Works in masonry were added to
the motte and the bailey banks in the 12th, 13th, and 14th centuries.
There are traces of a semicircular earthwork outside the second ditch
on the west, which appears to have formed a barbican. But the most
exceptional thing about this castle is the series of earthen platforms
on the north and east, connected by a bank, and closely investing
the external ditch, which were formerly supposed to form part of the
castle works. Mr W. St John Hope has suggested the far more plausible
theory that they were the siege platforms erected by Louis, the Dauphin
of France, in 1216. We are told that his engines kept up a most
destructive fire of stones.[284]

The value of the manor of Berkhampstead had considerably decreased,
even since the Count of Mortain received it.[285]

BISHOP'S STORTFORD, Herts (Fig. 9).--Waytemore Castle is the name given
to the large oval motte at this place, which is evidently the site of
the castle of "Estorteford," given by William the Conqueror to Maurice,
Bishop of London.[286] The manor of Stortford had been bought from King
William by Maurice's predecessor, William, who had been one of the
Norman favourites of Edward the Confessor.[287] He may have built this
castle, but he cannot have built it till after the Conquest, as the
land did not belong to his see till then.

"The castle consists of a large oval motte, 250 × 200 feet at its base,
rising 40 feet above the marshes of the river Stort, and crowned by
a keep with walls of flint rubble, 12 feet thick. On the S. of the
motte there are traces of a pentagonal bailey, covering 2-1/2 acres.
It is enclosed on four sides by the narrow streams which intersect
the marshes. The dry ditch on the fifth side, facing the motte, is
discernible. The castle abuts on the road called The Causeway, which
crosses the valley; it is in a good position to command both road and
river."[288] The value of the manor had gone down at Domesday.[289]

BOURN, Lincolnshire (Fig. 10).--The manor of Bourn or Brune appears to
have been much split up amongst various owners at the time of Domesday.
A Breton named Oger held the demesne.[290] A charter of Picot, the
Sheriff of Cambridgeshire, a person often mentioned in Domesday Book,
gives the church of Brune and the chapel of the castle to the priory
which he had founded near the castle of Cambridge--afterwards removed
to Barnwell.[291] Bourn was the centre of a large soke in Anglo-Saxon
times. Leland mentions the "Grete Diches, and the Dungeon Hill of the
ancient Castel,"[292] but very little of the remains is now visible,
and the motte has been almost removed.

"The castle lies in flat ground, well watered by springs and streams.
The motte was placed at the southern apex of a roughly oval bailey,
from which it was separated by its own wet ditch, access being obtained
through a gatehouse which stood on the narrow neck by which this
innermost enclosure, at its N.W. end, joined the principal bailey,
which, in its turn, was embraced on all sides but the S. by a second
and concentric bailey, also defended by a wet ditch, which broadens out
at the S.W. corner into St Peter's Pool. There is another enclosure
beyond this which may be of later date. The inner bailey covers 3
acres. Very little is now left of the motte, but a plan made in 1861
showed it to be fairly perfect,[293] and some slight remains of the
gatehouse were excavated in that year. The castle is on the line of
the Roman road from Peterborough to Sleaford, and close to the Roman
Car-Dyke."[294]

The value of Bourn had risen at Domesday.

[Illustration: FIG. 10. Bourn, Lincs. Bramber, Sussex.]

BRAMBER, Sussex (Fig. 10).--Of the manor of Washington, in which
Bramber is situated, the Survey says that it formerly paid geld
for fifty-nine hides; and in one of these hides sits the castle of
Bramber.[295] It must not be imagined that the castle _occupied_ a
whole hide, which according to the latest computations would average
about 120 acres. It is evident that there had been some special
arrangement between the King and William de Braose, the Norman
tenant-in-chief, by which the whole geld of the manor had been
remitted. The Domesday scribe waxes almost pathetic over the loss to
the fisc of this valuable prey. "It used to be ad firmam for 100_l._,"
he says. The manor of Washington belonged to Gurth, the brother of
Harold, before the Conquest, but it is clear that Bramber was not the
_caput_ of the manor in Saxon times; nor was Washington the centre of a
large soke. Bramber Castle was constructed to defend the estuary of the
river, now known as the Adur, one of the waterways to Normandy already
alluded to.

The castle occupies a natural hill which forms on the top a pear-shaped
area of 3 acres. Towards the middle rises an artificial motte about 30
feet high; there is no sign of a special ditch around it, except that
the ground sinks slightly at its base. The bailey is surrounded by a
very neatly built wall of pebbles and flints, laid herring-bone-wise in
places, which does not stand on an earthen bank. The absence of this
bank makes it likely, though of course not certain, that this wall was
the original work of De Braose; the stones of which it is composed
would be almost as easily obtained as the earth for a bank. On the
line of the wall, just east of the entrance, stands a tall fragment of
an early Norman tower. The workmanship of this tower, which is also
of flints laid herring-bone-wise, with quoins of ashlar, so strongly
resembles that of the neighbouring church that it seems obvious that
both were built at about the same time.[296] The church is dedicated to
St Nicholas, who was worshipped in Normandy as early as 1067;[297] it
was probably the Normans who introduced his worship into England. Both
church and tower are undoubtedly early Norman. The motte shows no sign
of masonry.

The value of the manor of Washington had slightly risen since the
Conquest.

BRISTOL.--Robert, Earl of Gloucester, the Empress Matilda's half-brother
and great champion, is always credited with the building of Bristol
Castle; but this is one of the many instances in which the man who
first rebuilds a castle in stone receives the credit of being the
original founder.[298] For it is certain that there was a castle at
Bristol long before the days of Earl Robert, as the _Anglo-Saxon
Chronicle_ mentions it in 1088, when it was held by Geoffrey, Bishop
of Coutances, and Robert Curthose against William II.; and Symeon of
Durham, in the same year, speaks of it as a "castrum fortissimum."
Bishop Geoffrey held Bristol at the date of the Domesday Survey,
and he probably built the castle by William's orders.[299] It was
completely destroyed in 1655 (only a few 13th century arches in a
private house now remain), and no trustworthy plan has been preserved,
but there is clear evidence that it was a motte-and-bailey castle of
the usual Norman type.[300] In Stephen's reign it was described as
standing on a very great _agger_.[301] An _agger_ does not necessarily
mean a motte, but it is often used for one, and there is other
evidence which shows that this is its meaning here. A Perambulation
of the bounds of Bristol in 1373 shows that the south-western part
of the castle ditch, which enclosed the site of the keep, was called
_le Mot-dich_; which should certainly be translated the ditch of the
motte, and not, as Seyer translates it, the moat ditch.[302] Finally,
the description of the castle in 1642 by Major Wood, says: "The castle
stood upon a lofty steep mount, that was not minable, as Lieutenant
Clifton informed me, for he said the mount whereon the castle stood
was of an earthy substance for a certain depth, but below that a firm
strong rock, and that he had searched purposely with an auger and found
it so in all parts."[303] He goes on to describe the wall of the bailey
as resting on an earthen rampart, testifying to the wooden stockade of
the first castle. The great tower of Earl Robert appears to have been
placed on the motte, which must have been of considerable size, as it
held not only the keep, but a courtyard, a chapel, and the constable's
house, besides several towers on its walls. The whole area of the
castle was very nearly 4 acres.[304]

Bristol Castle was no doubt originally a royal castle, though Earl
Robert of Gloucester held it in right of his wife, who had inherited it
from her father, Robert Fitz Hamon; but the crown did not abdicate its
claim upon it, and after the troubles of 1174, Henry II. caused the son
of Earl Robert to surrender the keep into his hands.[305]

Seyer very pertinently remarks that Bristol Castle "was erected with a
design hostile to the town; for it occupies the peninsula between two
rivers, along which was the direct and original communication between
the town and the main part of Gloucestershire."[306] It was outside the
city, and was not under its jurisdiction till James I. granted this
authority by charter.[307] The value T. R. E. is not given in Domesday
Book.

BUCKINGHAM.--The only mention of this castle as existing in the 11th
century is in the _Gesta Herewardi_,[308] an undated work which is
certainly in great part a romance, but as it is written by some one
who evidently had local knowledge, we may probably trust him for the
existence of Buckingham Castle at that date; especially as Buckingham
was a county town, and one of the boroughs of the Burghal Hidage, the
very place which we should expect to find occupied by a Norman castle.
This writer speaks of the castle as belonging to Ivo de Taillebois;
this is not inconsistent with the fact shown by Domesday Book, that
the borough belonged to the king. That it was a motte-and-bailey
castle is indicated by Speed's map of Buckingham in 1611; he speaks
of the "high hill," though he only indicates it slightly in his plan,
with a shield-shaped bailey. Brayley states that the present church
is "proudly exalted on the summit of an artificial mount, anciently
occupied by a castle."[309]

The castle hill occupies a strong position on the neck of land made
by a bend of the river; it extends nearly half-way across it, and
commands both town and river. The original earthworks of the castle
were destroyed and levelled for the erection of a church in 1777, but
the large oval hill remains, having a flat summit about 2 acres in
extent, and about 30 feet above the town below. Its sides descend in
steep scarps behind the houses on all sides but the north-east. There
can be no doubt that the motte has been lowered, and thus enlarged, in
order to build the church. The foundations of a stone castle were found
in digging a cellar on the slope of the motte.[310]

The value of Buckingham had considerably risen at the date of
Domesday.[311]

CAERLEON, Monmouthshire (Fig. 11).--Domesday Book speaks of the
_castellaria_ of Caerleon.[312] A _castellaria_ appears to have meant
a district in which the land was held by the service of castle-guard
in a neighbouring castle. The Survey goes on to say that this land was
waste in the time of King Edward, and when William de Scohies, the
Domesday tenant, received it; _now_ it is worth 40s. _Wasta_, Mr Round
has remarked, is one of the pitfalls of the Survey. Perhaps we shall
not be far wrong if we say that in a general way it means that there
was nobody there to pay geld. When this occurs in a town it may point
to the devastations committed at the Conquest; but when it occurs in
the country, and when it is accompanied by so clear a statement that
the land which was _wasta_ in King Edward's time and at the Conquest is
now producing revenue, the inference would seem to be clear that the
castle of Caerleon was built on uninhabited land. Caerleon, however,
had been a great city in Roman times, and had kept up its importance
at least till the days of Edgar, when it is twice mentioned in Welsh
history.[313] It must therefore have gone downhill very rapidly.
Giraldus mentions among the ruins of Roman greatness which were to be
seen in his day, a gigantic tower, and this is commonly supposed to
have belonged to the castle.[314] It certainly did not, for Giraldus is
clearly speaking of a Roman tower, and the motte of the Norman castle
not only has no signs of masonry, but has been thrown up over the ruins
of a Roman villa which had been burnt.[315] The motte and other remains
of the castle are outside the Roman _castrum_, between it and the
river. The bailey is roughly pentagonal, and covers 4-3/4 acres. The
manor of Caerleon was waste T. R. E. and had risen to 40s. T. R. W.[316]

[Illustration: FIG. 11. Caerleon, Monmouth. Carisbrooke.]

CAMBRIDGE.--Ordericus tells us that William built this castle on his
return from his first visit to Yorkshire in 1068,[317] and Domesday
Book states that twenty-seven houses were destroyed to make room for
the castle.[318] There can hardly be a clearer statement that the
castle was entirely new. We have already seen that there is some
probability that Cambridge was first fortified by the Danes; for though
it has been assumed to be a Roman castrum, no Roman remains have
ever been found there, and the names which suggest Roman occupation,
Chesterton and Grantchester, are at some distance from Cambridge. The
castle, according to Mr St John Hope's plan,[319] was placed inside
this enclosure, and the destruction of the houses to make room for it
is thus explained. The motte and a portion of the bank of the bailey
are all that now remain of the castle, but the valuable ancient maps
republished by Mr Hope show that the motte had its own ditch, and
that the bailey was rectangular. There was formerly a round tower on
the motte, which, if it had the cross-loop-holes and machicolations
represented in the print published in 1575, was certainly not of Norman
date. The area of the bailey was 4-1/4 acres.[320] The castle was a
royal one, and like many royal castles, went early to ruin. Henry IV.
gave the materials of the hall to the master and wardens of King's Hall
for building their chapel.

The value of Cambridge T. R. W. is not given in Domesday Book.

CANTERBURY.--Domesday Book only mentions this castle incidentally
in connection with an exchange of land: "The archbishop has seven
houses and the abbot of St Augustine fourteen for the exchange of
the castle."[321] It has been too hastily assumed that it was a
pre-Conquest castle which was thus exchanged for twenty-one houses;
but anyone who knows the kind of relations which existed chronically
between the archbishop of Canterbury and the abbot of St Augustine's
will perceive that it was an impossibility that these two potentates
should have held a castle in common. It was the land for the castle,
not the castle itself, which the king got from these ecclesiastics.
This is rendered clear by a passage in the Chartulary of St Augustine's,
which tells us that the king, who was mesne lord of the city of
Canterbury, had lost the rent of thirty-two houses through the exchange
of the castle: seven having gone to the archbishop, fourteen to the
abbot, and eleven having been destroyed in making the ditch of the
castle.[322] There can scarcely be any doubt that the hillock now known
by the ridiculous name of Dane John is the motte of this original
castle of the Conqueror. Its proper name, the Dungeon Hill, which it
bore till the 16th and even the 18th century,[323] shows what its
origin was; it was the hill on which stood the dungeon or donjon of
a Norman castle.[324] The name Dane John is not so much a corruption
as a deliberate perversion introduced by the antiquary Somner about
1640, under the idea that the Danes threw up the hill--an idea for
which there is not the slightest historical evidence.[325] We have
seen that there is no reason to think that the Danes ever constructed
fortifications of this kind, and their connection with this earthwork
is due to one of those guesses, too common in English archæology, which
have no scientific basis whatever.

Somner makes the important statement that this earthwork was originally
outside the city walls. His words are:--

    "I am persuaded (and so may easily, I think, anyone be that well
    observes the place) that the works both within and without the
    present wall of the city were not counterworks one against the
    other, as the vulgar opinion goes, but were sometimes all one
    entire plot containing about 3 acres of ground, of a triangular
    form (the outwork) with a mount or hill entrenched round within it;
    and that when first made or cast up it lay wholly without the city
    wall; and hath been (the hill or mount, and most part also of the
    outwork), for the city's more security, taken in and walled since;
    that side of the trench encompassing the mound now lying without
    and under the wall fitly meeting with the rest of the city ditch,
    after either side of the earthwork was cut through to make way for
    it, at the time of the city's inditching."[326]

It is not often we are so fortunate as to have so clear a description
of an earthwork which has almost entirely disappeared; but the
description is confirmed by Stukeley and Hasted, and down to the making
of the Chatham and Dover railway in 1860 the earthworks of the part of
the bailey which was left outside the city wall were still to be seen,
and were noticed by Mr G. T. Clark.[327] It is clear that Somner's
description corresponds exactly, even in the detail of size, to the
type of a motte-and-bailey castle.

There are certain facts, which have not been put together before,
which enable us to make a very probable guess as to the date at which
this ancient castle was cut through by the newer city bank. The walls
of Canterbury have never yet received so careful an examination as
those of Rochester have had from the Rev. Greville Livett;[328] but
the researches of Mr Pilbrow about thirty years ago showed that the
original Roman walls included a very small area, which would leave
both the motte and the Plantagenet castle outside.[329] Certain
entries in the _Close Rolls_ show that the fortification of the town
of Canterbury was going on in the years 1215-1225.[330] But it is
too often forgotten that where a wall stands on an earthen bank it
is a clear proof that before the wall was built there was a wooden
stockade in its place. Now the portion of the city wall which encloses
the Dane John stands on an earthen bank; so, indeed, does the whole
wall from the Northgate to the castle. It is clear that this piece of
bank cannot have been made till the first Norman castle, represented
by the earthwork, was abandoned; and fortunately we have some evidence
which suggests a date for the change. In the _Pipe Rolls_ of Henry
II.'s reign there are yearly entries, beginning in 1168, of 5s. paid to
Adeliza Fitzsimon "for the exchange of her land which is in the castle
of Canterbury." There can be little doubt that this land was purchased
to build the great Plantagenet castle whose splendid keep was once one
of the finest in England.[331] The portion of the castle wall which
can still be seen does not stand on an earthen bank, an indication
(though not a proof) that the castle was on a new site. Henry II. was a
great builder of stone keeps, but he seldom placed them on artificial
mottes. It is no uncommon thing to find an old motte-and-bailey castle
abandoned for a better or larger site close at hand.[332]

The bailey of the second castle, according to Hasted, extended almost
to the Dane John, which is about 800 feet from the present keep. The
part of the older castle which lay outside the new city bank was
possessed by a family of the name of Chiche from the time of Henry
II. to that of Edward IV., while the Dungeon Hill itself remained
royal property.[333] That the new bank was Henry II.'s work we may
conjecture from the passages in the _Pipe Rolls_, which show that
between the years 1166 and 1173 he spent about £30 in enclosing the
city of Canterbury and making a gate. We are therefore not without
grounds for concluding that Henry II. was the first to enlarge the city
by taking in the Dane John, cutting through the ancient bailey, and at
the same time enclosing a piece of land for a new stone castle.[334]
The very small sum paid for the city gate (11s., equal to about £11
of our money) suggests that the gate put up by Henry II. was a wooden
gateway in the new stockaded bank. The stone walls and towers which
were afterwards placed on the bank are of much later date than his
reign.[335]

The Dungeon Hill appears to have been used for the last time as a
fortification in 1643, when ordnance was placed upon it, and it
was ordered to be guarded by the householders.[336] In 1790 it was
converted into a pleasure-ground for the city; the wide and deep ditch
which had surrounded it was filled up, and serpentine walks cut to
lead up to the summit. Brayley says that "the ancient and venerable
character of this eminence was wholly destroyed by incongruous
additions." Still, enough remains to show that it was once a very fine
motte, such as we might expect the Conqueror to raise to hold in check
one of the most important cities of his new realm.

The value of Canterbury had increased from 51_l._ to 54_l._ since the
days of King Edward.[337]

CARISBROOKE, Isle of Wight (Fig. 11).--There can be no doubt that
this is the castle spoken of in Domesday Book under the manor of
Alwinestone. Carisbrooke is in the immediate neighbourhood of
Alvington. The language in which the Survey speaks of this manor is
worthy of note. "The king holds Alwinestone: Donnus held it. It then
paid geld as two and a half hides: now as two hides, because the castle
sits in one virgate."[338] Certain entries similar to this in other
places seem to indicate that there was some remission of geld granted
on the building of a castle;[339] but as here the king was himself the
owner, the remission must have been granted to his tenants.

The original castle of Carisbrooke consists of a high motte, ditched
round, placed at the corner of a parallelogram with rounded corners.
This bailey, covering 2-3/4 acres, is surrounded by high banks, which
testify to the former presence of a wooden stockade. There is another
bailey on the eastern side, called the Tilt-yard. The excellent little
local guide-book compiled by Mr Stone calls this a British camp, but
there is no reason to believe that it was anything else than what it
appears to be--a second bailey added as the castle grew in importance.
On the motte is a shell of polygonal form, of rubble masonry, but
having quoins of well-dressed ashlar. It is believed to be of the time
of Henry I., since the author of the _Gesta Stephani_ states that
Baldwin de Redvers, son of Richard de Redvers, to whom Henry granted
the lordship of the Isle of Wight, had a castle there splendidly
built of stone, defended by a strong fortification.[340] This would
indicate that, besides the stone keep, stone walls were added to the
earthworks of the Domesday castle. The keep is of peculiar interest,
as it still retains the remains of the old arrangements in keeps of
this style, though of much later date. The motte was opened in 1893,
and was found to be composed of alternate layers of large and small
chalk rubble.[341] Little attention has hitherto been paid to the
construction of these Norman mottes, but other instances have been
noted which show that they were often built with great care. The whole
castle, including the Tilt-yard, was surrounded with an elaborate
polygonal fortification in Elizabeth's reign, when the Spanish invasion
was expected.

The value of the manor of Alvington had increased at the time of the
Survey, though the number of ploughs employed had actually decreased.
This increase must have been owing to the erection of the castle, which
provided security for trade and agriculture. Alvington was not the
centre of a large soke in the Confessor's time, so it is unlikely that
there was any fortification there in Saxon days.[342]

CARLISLE, Cumberland (FIG. 12).--This castle was built by William Rufus
in 1092, when for the first time Cumberland was brought under Norman
sway. The _Anglo-Saxon Chronicle_ says, "he repaired the _burh_, and
reared the castle," a passage which is sufficient of itself to show
that _burh_ and castle were two quite different things. Carlisle of
course was a Roman fortress, and needed only the repairing of its
walls. The castle was a new thing, and was placed outside the city. Its
plan, which is roughly a triangle, with the apex formed into a small
court by a ditch which (formerly) separated it from the bailey, looks
very suggestive of a previous motte and bailey, such as we might expect
the Norman king to have thrown up. The keep is known to have been built
by David, king of Scotland, in Stephen's reign,[343] and it is possible
that he may have removed the motte. The castle appears to have had a
wooden _pelum_ or _palicium_ on its outer banks as late as 1319.[344]
The whole area covers 4 acres.

CASTLE ACRE, Norfolk (Fig. 12).--There can be no doubt that this castle
existed in the 11th century, as William de Warenne mentions it in the
charter of foundation of Lewes Priory, one of the most interesting and
human of monastic charters.[345] The earthworks still remaining of
this castle are perhaps the finest castle earthworks in England; the
banks enclosing the bailey are vast. The large and high motte carries
a wall of flint rubble, built outside and thus revetting the earthen
bank which formed its first defence. In the small court thus enclosed
(about 100 feet in diameter) the foundations of an oblong keep can
be discerned. A very wide ditch surrounds the motte, and below it is
a horse-shoe bailey, about 2 acres in extent, stretching down to the
former swamps of the river Nar. On the east side of the motte is a
small half-moon annexe, with its own ditch; this curious addition is
to be found in several other motte castles,[346] and is believed to
have been a work intended to defend the approach, of the nature of
a barbican. On the west side of the motte is the village of Castle
Acre, enclosed in an oblong earthwork with an area of 10 acres. This
work now goes by the name of the Barbican, but probably this name
has been extended to it from a barbican covering the castle entrance
(of which entrance the ruins still remain). It is most likely that
this enclosure was a _burgus_ attached to the castle. Mr Harrod, who
excavated the banks, found quantities of Roman pottery, which led him
to think that the work was Roman; but as the pottery was all broken,
it is more likely that the banks were thrown up on the site of some
Roman villa.[347] This earthwork has a northern entrance in masonry,
evidently of 13th century date; and as the scanty masonry remaining
of the castle is similar in character, it is probably all of the same
date. The area covered by the motte and the two original baileys is
3-1/2 acres; that of the whole series of earthworks, 15 acres.

Acre was only a small manor in Saxon times; its value at the time of
the Survey had risen from 5_l._ to 9_l._[348]

[Illustration: FIG. 12. Carlisle. Castle Acre, Norfolk.]

CHEPSTOW (Estrighoel or Strigul), Monmouthshire.--Notwithstanding the
fact that there is another castle of the name of Strigul about 9 miles
from Chepstow (known also as Troggy Castle), it is clear that Chepstow
is the castle meant by Domesday, as the entry speaks of ships going
up the river, a thing impossible at Strigul.[349] The castle occupies
a narrow ridge, well defended by the river on one side, and on the
other by a valley which separates it from the town. There are four
wards, and the last and smallest of all seemed to the writer, when
visiting the castle, to mark the site of a lowered motte. This opinion,
however, is not shared by two competent observers, Mr Harold Sands and
Mr Duncan Montgomerie, who had much ampler opportunities for studying
the remains. This ward is now a barbican, and the masonry upon it
belongs clearly to the 13th century; it occupies the highest ground in
the castle, and is separated from the other wards, and from the ridge
beyond it, by two ditches cut across the headland. The adjoining court
must have belonged to the earliest part of the castle, as it contains
a very remarkable early Norman building (splendidly restored in the
13th century) which is regarded by most authorities as the original
hall of William FitzOsbern. It must, however, have combined both hall
and keep, otherwise the castle was not provided with any citadel, if
there was no motte.[350] What is now the second ward has a Norman
postern in the south wall, and may have been the bailey to the keep.
All the other masonry is of the late Early English or the Perpendicular
period, and the entrance ward is probably an addition of the 13th
century. The shape of all the baileys is roughly quadrangular, except
that of the fourth, which would be semicircular but for the towers
which make corners to it. The whole area of the castle is 1-2/3 acres.

We are not told what the value of the manor was before William
FitzOsbern built his castle there, but from the absence of this mention
we may infer that the site was waste. It paid 40_s._ in his time from
ships' dues, 16_l._ in his son Earl Roger's time, and at the date of
the Survey it paid the king 12_l._[351] Chepstow was not the centre of
a large soke, and it appears to have owed all its importance to the
creation of William FitzOsbern's castle.

CHESTER.--The statement of Ordericus, that William I. founded this
castle on his return from his third visit to York, is sufficiently
clear.[352] The very valuable paper of Mr E. W. Cox on Chester
Castle[353] answers most of the questions which pertain to our present
inquiry. The original castle of Chester consisted of the motte, which
still remains, though much built over, and the small ward on the
edge of which it stands, a polygonal enclosure scarcely an acre in
extent. On the motte the vaulted basement of a tower still remains,
but the style is so obscured by whitewash and modern accretions that
it is impossible to say whether the vaulting is not modern. The first
buildings were certainly of wood, but Mr Cox regarded some of the
existing masonry on the motte as belonging to the 12th century; and
this would correspond with the entry in the _Pipe Rolls_ of 102_l._
7_s._ 0_d._ spent on the castle by Henry II. in 1159.[354] The tower,
nicknamed Cæsar's Tower, and frequently mistaken for the keep, is
shown in Mr Cox's paper to be only a mural tower of the 13th century,
probably built when the first ward was surrounded with walls and towers
in masonry.[355] The large outer bailey was first added in the reign
of Henry III.[356] It is further proved by Mr Cox that Chester Castle
stood outside the walls of the Roman city. The manor of Gloverstone lay
between it and the city, and was not under the jurisdiction of the city
until quite recent times.[357] This disposes of the ball set rolling by
Brompton at the end of the 13th century, and sent on by most Chester
topographers ever since, that Ethelfleda, when she restored the Roman
walls of Chester, enlarged their circuit so as to take in the castle.
We have already referred to this in Chapter III.

Chester, as we have seen, was originally a royal castle. And though it
was naturally committed to the keeping of the Norman earls of Chester,
and under weak kings may have been regarded by the earls as their own
property, no such claim was allowed under a strong ruler. After the
insurrection of the younger Henry, Hugh, Earl of Chester, forfeited his
lands; Henry II. restored them to him in 1177, but was careful to keep
the castle in his own hands.[358]

The city of Chester, Domesday Book tells us, had greatly gone down in
value when the earl received it, probably in 1070; twenty-five houses
had been destroyed. But it had already recovered its prosperity at the
date of the Survey; there were as many houses as before, and the ferm
of the city was now let by the earl at a sum greatly exceeding the ferm
paid in King Edward's time.[359] This prosperity must have been due to
the security provided for the trade of Chester by the Norman castle and
Norman rule.

[Illustration: FIG. 13. Clifford, Hereford. Clitheroe, Lancs. Corfe,
Dorset.]

CLIFFORD, Herefordshire (Fig. 13).--It is clearly stated by Domesday
Book that William FitzOsbern built this castle on waste land.[360]
At the date of the Survey it was held by Ralph de Todeni, who had
sub-let it to the sheriff. In the many castles attributed to William
FitzOsbern, who built them as the king's vicegerent, we may see an
indication that the building of castles, even on the marches of Wales,
was not undertaken without royal license. In the reign of Henry I.
Clifford Castle had already passed into the hands of Richard Fitz
Pons, the ancestor of the celebrated house of Clifford, and one of the
_barons_ of Bernard de Neufmarché, the Norman conqueror of Brecon.[361]

The castle has a large motte, roughly square in shape, which must be
in part artificial.[362] Attached to it on the south-west is a curious
triangular ward, included in the ditch which surrounds the motte.
The masonry on the motte is entirely of the "Edwardian" style, when
keepless castles were built; it consists of the remains of a hall, and
a mural tower which is too small to be called a keep. There is also a
small court, with a wall which stands on a low bank. Below the motte is
an irregular bailey of about 2-1/3 acres, with earthen banks which do
not appear to have ever carried any masonry, though in the middle of
the court there is a small mound which evidently covers the remains of
buildings. The whole area of the castle, including the motte and the
two baileys, is about 3-1/2 acres.

The value of the manor had apparently risen from nothing to 8_l._ 5_s._
Clifford was not the centre of a large soke.

CLITHEROE, Lancashire (Fig. 13).--There is no express mention of this
castle in Domesday Book, but of two places in Yorkshire, Barnoldswick
and Calton, it is said that they are in the _castellate_ of Roger the
Poitevin.[363] A castellate implies a castle, and as there is no other
castle in the Craven district (to which the words of the Survey relate)
except Skipton, which did not form part of Roger's property, there is
no reason to doubt that this castle was Clitheroe, which for centuries
was the centre of the Honour of that name. The whole land between the
Ribble and the Mersey had been given by William I. to this Roger, the
third son of his trusted supporter, Earl Roger of Shrewsbury. One can
understand why William gave important frontier posts to the energetic
and unscrupulous young men of the house of Montgomeri, one of whom was
the adviser and architect of William Rufus, another a notable warrior
in North Wales, another the conqueror of Pembrokeshire. As it appears
from the Survey that Roger's possessions stretched far beyond the
Ribble into Yorkshire and Cumberland, it seems quite possible--though
here we are in the region of conjecture--that just as his father and
brothers had a free hand to conquer as they listed from the North and
South Welsh, so Roger had a similar commission for the hilly districts
still unconquered in the north-west of England. But fortune did not
favour the Montgomeri family for long. They were exiled from England in
1102 for siding with Robert Curthose, and in the same year we find the
castle of Clitheroe in the hands of Robert de Lacy, lord of the great
Yorkshire fief of Pontefract.[364]

The castle of Clitheroe stands on a lofty motte of natural rock.[365]
There are no earthworks on the summit, but a stout wall of limestone
rubble without buttresses encloses a small court, on whose south-west
side stands the keep. It is just possible that the outer wall may be
the original work of Roger, as limestone rubble would be easier to
get than earth on this rocky hill. The keep is small, rudely built of
rubble, and has neither fireplace nor garde-robe, nor the slightest
ornamental detail--not even a string course. But in spite of the entire
absence of ornament, a decorative effect has been sought and obtained
by making the quoins, voussoirs, and lintels of a dressed yellow
sandstone. The care with which this has been done is inconsistent with
the haste with which Roger must inevitably have constructed his first
fortification, if we suppose, as is probable, that he received the
first grant of his northern lands on William's return in 1070 from his
third visit to the north, when he made that remarkable march through
Lancashire to Chester which is described by Ordericus. It seems more
likely that even if the outer wall or shell were the work of Roger, he
had only wooden buildings inside its circuit. Dugdale attributes the
building of the keep to the second Robert de Lacy, between 1187 and
1194, and it is probable that this date is correct.[366] The bailey of
Clitheroe lay considerably below the keep, and is now overbuilt with a
modern house, offices, and garden. It covers one acre. A Roman road up
the valley of the Ribble passes near the foot of the rock.[367]

As the very name of Clitheroe is not mentioned in Domesday Book, it
clearly was not an important centre in Saxon times. The value of
Blackburn Hundred, in which Clitheroe is situated, had fallen between
the Confessor's time and the time when Roger received it. It is quite
possible that he never lived at Clitheroe, as he sub-infeoffed the
manor and Hundred of Blackburn to Roger de Busli and Albert Greslet
before 1086.[368]

COLCHESTER, Essex.--The remarkable keep of this castle has been the
subject of antiquarian legend for many centuries, and Mr Clark has
the merit of having proved its early Norman origin, by its plan and
architecture. A charter of Henry I. is preserved in the cartulary
of St John's Abbey at Colchester, which grants to Eudes the Dapifer
"the city of Colchester, and the tower and the castle, and all the
fortifications of the castle, just as my father had them and my
brother and myself."[369] This proves that the keep and castle were
in existence in the Conqueror's time; the Norman character of the
architecture proves that the keep was not in existence earlier. We see,
then, that the reason there is no motte at Colchester is that there was
a stone keep built when first the castle was founded. As far as we are
aware, Colchester, the Tower of London, and the recently discovered
keep of Pevensey are the only certain instances of stone keeps of the
11th century in England.

That one of the most important of the Conqueror's castles, second
only to the Tower of London, and actually exceeding it in the
area it covers, should be found in Colchester, is not surprising,
because the Eastern counties at the time of the Conquest were not
only the wealthiest part of the kingdom (as Domesday Book clearly
shows[370]), but they also needed special protection from the attacks
of Scandinavian enemies. Mr Round has conjectured that the castle
was built at the time of the invasion of St Cnut, between 1080 and
1085.[371]

The castle is built of Roman stones used over again, with rows of tiles
introduced between the courses with much decorative effect.[372] The
original doorway was on the first floor, as in most Norman keeps; but
at some after time, probably in the reign of Henry I.,[373] the present
doorway was inserted; and most likely the handsome stairway which now
leads up from this basement entrance was added, as it shows clear marks
of insertion. Henry II. was working on the walls of the castle in 1182,
and it may be strongly suspected that the repairs in ashlar, and the
casing of the buttresses with ashlar, were his work.[374] One item
in the accounts of Henry II. is £50 "for making the bailey round the
castle."[375] There were two baileys to the castle of Colchester--the
inner one, which scarcely covered 2 acres, and the outer one, which
contained about 11. The inner bailey was enclosed at first with an
earthwork and stockade, the earthwork being thrown up over the remains
of some Roman walls, whose line it does not follow. Afterwards a stone
wall was built on the earthwork, the foundations of which can still be
traced in the west rampart.[376] The outer bailey, which lay to the
north, extended on two sides to the Roman walls of the town; on the
west side it had a rampart and stockade. If the £50 spent by Henry II.
represents the cost of a stone wall round the inner bailey, then the
_palicium_ blown down by the wind in 1219 must have been the wooden
stockade on the west side of the outer bailey.[377] The question is
difficult to decide, but at any rate the entry proves that as late
as Henry III.'s reign, some part of the outer defences of Colchester
Castle was still of timber.

The position of Colchester Castle is exceptional in one respect, that
the castle is almost in the middle of the town. But this very unusual
position is explained by Mr Round's statement that the land forming the
castle baileys, as well as that afterwards given to the Grey Friars
on the east, was crown demesne before the Conquest, and consequently
had been cultivated land, so that we do not hear of any houses in
Colchester being destroyed for the site of the castle.[378] But by
keeping this land as the inalienable appendage of the royal castle
William secured that communication between the castle and the outside
country which was so essential to the invaders.

The value of the city of Colchester had risen enormously at the date of
the Survey.[379]

CORFE, Dorset (Fig. 13).--Mr Eyton has shown that for the _castellum
Warham_ of Domesday Book we ought to read _Corfe_, because the castle
was built in the manor of Kingston, four miles from Wareham.[380]
And this is made clear by the _Testa de Nevill_, which says that
the church of Gillingham was given to the nunnery of Shaftesbury in
exchange for the land on which the castle of Corfe is placed.[381]
Because King Edward the Martyr was murdered at Corfe, at some place
where his stepmother Elfrida was residing, it has been inferred that
there was a Saxon castle at Corfe; and because there is a building with
some herring-bone work among the present ruins, it has been assumed
that this building is the remains of that castle or palace. But the
_Anglo-Saxon Chronicle_, the only contemporary authority for the event,
says nothing of any castle at Corfe, but simply tells us that Edward
was slain at Corfe Geat, a name which evidently alludes to a gap or
passage through the chalk hills, such as there is at Corfe.[382] Nor
is there any mention of Corfe as a fortress in Anglo-Saxon times;
it is not named in the _Burghal Hidage_, and we do not hear of any
sieges of it by the Danes. Nor is it likely that the Saxons would
have had a fortress at Corfe, when they had a fortified town so near
as Wareham.[383] Kingston, the manor in which Corfe is situated, was
not an important place, as it had no dependent soke. The language of
Domesday absolutely upsets the idea of any Saxon castle or palace at
Corfe, as it tells us that William obtained the land for his castle
from the nuns of Shaftesbury, and we may be quite sure they had no
castle there.[384]

Corfe Castle stands on a natural hill, which has been so scarped
artificially that the highest part now forms a large motte. Three wards
exist--the eastern or motte ward, the western, and the southern. The
two former probably formed the original castle. On the motte (which
possibly is not artificial, but formed by scarping) stands the lofty
keep, of splendid workmanship, probably of the time of Henry I. In
the ward pertaining to it are buildings of the time of John and Henry
III.[385] The western ward has towers of the 13th century, but it also
contains the interesting remains of an early Norman building, probably
a hall or chapel, built largely of herring-bone work; this is the
building which has been so positively asserted to be a Saxon palace.
But herring-bone masonry, which used to be thought an infallible sign
of Saxon work, is now found to be more often Norman.[386] The building
is certainly an ancient one, and may possibly have been contemporary
with the first Norman castle; its details are unmistakably Norman.
But very likely it was the only Norman masonry of the 11th century at
Corfe Castle.[387] It is clear that the stone wall which at present
surrounds the western bailey did not exist when the hall (or chapel)
was built, as it blocks up its southern windows. Probably there was a
palisade at first on the edge of the scarp. Palisades still formed part
of the defences of the castle in the time of Henry III., when 62_l._
was paid "for making two good walls in place of the palisades at Corfe
between the old bailey of the said castle and the middle bailey towards
the west, and between the keep of the said castle and the outer bailey
towards the south."[388] This shows that the present wing-walls down
from the motte were previously represented by stockades. The ditch
between the keep and the southern bailey has been attributed to King
John, on the strength of an entry in the _Close Rolls_ which orders
fifteen miners and stone-masons to work on the banks of the ditch in
1214.[389] But we may be quite certain that this ditch below the motte
belonged to the original plan of the castle; John's work would be
either to line it with masonry, or to enlarge it. It is not without
significance for the early history of the castle that Durandus the
_carpenter_ held the manor of Mouldham near Corfe, by the service of
finding a carpenter to work at the keep whenever required.[390]

The area of Corfe Castle, if we include the large southern bailey,
is 3-3/4 acres; without it, 1-1/2 acres. This bailey was certainly in
existence in the reign of Henry III. (as the extract from the _Close
Rolls_ proves) before the towers of superb masonry were added to it by
Edward I.

The value of Kingston Manor had considerably increased at the date of
the Survey. After the Count of Mortain forfeited his lands (in 1105),
the castle of Corfe was kept in the hands of the crown, and this
increases the probability that the keep was built by Henry I.

About 400 yards S.W. of Corfe Castle is an earthwork which might be
called a "Ring and Bailey." Instead of the usual motte there is a
circular enclosure, defended by a bank and ditch of about the same
height as those of its bailey, but having in addition an interior
platform or berm. This work is probably the remains of a camp thrown up
by Stephen during his unsuccessful siege of Corfe Castle in 1139.

[Illustration: FIG. 14. Dover. (From a plan in the British Museum,
1756.)]

DOVER, Kent (Fig. 14).--The Norman historian, William of Poitiers,
tells us that the castrum of Dover was built by Harold at his own
expense.[391] This comes from the celebrated story of the oath of
Harold to William, a story of which Mr Freeman says that there is no
portion of our history more entangled in the mazes of contradictory
and often impossible statements.[392] But let us assume the statement
about the _castrum_ to be true; the question then to be answered is
this: of what nature was that castrum? We never are told by English
chroniclers that Harold built any castles, though we do hear of his
fortifying towns. The present writer would answer this question,
tentatively indeed, and under correction, by the theory that the
castrum constructed or repaired by Harold was the present outer rampart
of Dover Castle, which encloses an area of about 34 acres, and may have
enclosed more, if it was formerly complete on the side towards the
sea.[393] The evidence in support of this theory is as follows:--

1. There certainly was a _burh_ on the top of the cliff at Dover in
Saxon times, as the _Anglo-Saxon Chronicle_ tells us that in 1048
Eustace of Boulogne, after coming to Dover, and slaying householders
there, went _up to the burh_, and slew people both within and without,
but was repulsed by the burh-men.[394] There was then a burh, and
valiant burh-men on the cliff at Dover in Edward the Confessor's reign.
But the whole analogy of the word burh makes it certain that by the
time of Edward it meant a fortified town.[395]

2. That the burh at Dover was of the nature of a town, with houses
in it, is confirmed by the poem of Guy of Amiens, who says that when
King William entered the _castrum_, he ordered the English to evacuate
their houses.[396] William of Poitiers also states that there was an
innumerable multitude of people in the castle,[397] though he may refer
to a multitude gathered there for safety.

3. Though the whole of the outer enceinte is generally credited to
Hubert de Burgh in Henry III.'s reign, the truth probably is that he
built the first stone walls and towers on the outer rampart; but the
existence of this earthen rampart shows that there was a wooden wall
upon it previously. It is not improbable that it was for the repair of
this wooden wall that so much timber was sent to Dover in the reigns
of Richard I. and John.[398] Dering, who was lieutenant of the castle
in 1629, records the tradition that the tower in the outer enceinte,
called Canons' Gate, dates from Saxon times (of course this could only
be true of a wooden predecessor of the stone tower), and that Godwin's
Tower, on the east side of the outer vallum, existed as a postern
before the Conquest.[399] Nearly all the towers on this wall were
supported by certain manors held on the tenure of castle-guard, and
eight of them still retain the names of eight knights to whom William
is said to have given lands on this tenure. Mr Round has shown that
the _Warda Constabularii_ of Dover Castle can be traced back to the
Conquest, and that it is a mere legend that it was given as a fief to a
Fienes. He remarks that the nine wards of the castle named in the Red
Book of the Exchequer are all reproduced in the names still attached to
the towers. "This coincidence of testimony leads us to believe that the
names must have been attached at a very early period; and looking at
the history of the families named, it cannot have been later than that
of Henry II."[400] May it not have been even earlier? Eight of these
names are attached to towers on the outer circuit,[401] and five of
them are found as landholders in Kent in Domesday Book.

4. William of Poitiers further tells us that when the duke had taken
the castle, he remained there eight days, _to add the fortifications
which were wanting_.[402] What was wanting to a Norman eye in
Anglo-Saxon fortifications, as far as we know them, was a citadel; and
without laying too much stress on the chronicler's eight days, we may
assume that the short time spent by William at Dover was just enough
for the construction of a motte and bailey, inside the _castrum_ of
Harold, but crowned by wooden buildings only.

Taking these things together, we venture to assume that the inner court
in which the keep of Dover stands, represents an original motte, or at
any rate an original citadel, added to the castle by William I. Whether
what now remains of this motte is in part artificial, we do not pretend
to say; it may be that it was formed simply by digging a deep ditch
round the highest knoll of ground within the ancient ramparts.[403]
Anyhow, it is still in effect a motte, and a large one, containing not
only the magnificent keep, but a small ward as well. That this keep was
the work of Henry II. there can be no manner of doubt; the _Pipe Rolls_
show that he spent more than £2000 on the _turris_ or keep of Dover
Castle between the years 1181 and 1187, and Benedict of Peterborough
mentions the building of the keep at this date.[404] The curtain around
the motte may also be reckoned to be his work originally, as the
_cingulum_ is spoken of along with the _turris_ in the accounts. Modern
alterations have left little of Norman character in this curtain which
shows at a glance, and the gateways (one of which remains) belong to a
later period.

Attached to this keep ward is another ward, whose rampart is generally
attributed to Saxon times. We are not in a position positively to
deny that the Saxons had an inner earthwork on the highest part of
the ground within their _burh_. But considering that small citadels
are unusual in Saxon earthworks: considering also that this bailey is
attached to the motte in the usual manner of a Norman bailey, and that
its size corresponds to the usual size of an original Norman bailey in
an important place, it does not seem unreasonable to suppose that this
was the original bailey attached to the Conqueror's motte. Its shape
is singular, part of it being nearly square, while at the S.E. corner
a large oval loop is thrown out, so as to enclose the Roman Pharos and
the Saxon church. The outline of the bailey certainly suggests that it
was built after the Pharos and the church, and was built with reference
primarily to the keep or motte ward. The nature of the ground, and
the necessity of enclosing the church and the Roman tower within
the immediate bailey of the castle, which would otherwise have been
commanded by them, were the other factors which decided the unusual
shape of the bailey.

On this earthwork the foundations of a rubble wall were formerly to
be traced,[405] probably built by Henry II., as considerable sums for
"the wall of the castle" are mentioned in his accounts.[406] Whether
there are still any remains of this curtain we are unable to say, but
so many of the features of the middle ward have been swept away by
modern alterations, and the difficulty of examining what remains, owing
to military restrictions, is so great, that little can be said about
it, and we find that most authorities observe a judicious silence on
the subject. But as the carriage of stone is expressly mentioned in
Henry II.'s accounts, we may with great probability assign to him the
transformation of the original wooden castle of William into a castle
of stone; while the transformation of the Anglo-Saxon borough into a
stone enceinte was the work of Henry III.'s reign.

We think the evidence suggests that this _burh_ or outer rampart was in
existence when the Conqueror came to Dover, crowned in all probability
with a stockade and towers of wood. It may possibly have been a British
or even a Roman earthwork originally (though its outline does not
suggest Roman work); or it may have been built by Harold as a city of
refuge for the inhabitants of the port.[407] The Saxon church which it
encloses, and which has long been attributed to the earliest days of
Saxon Christianity, is now pronounced by the best authorities to be
comparatively late in the style.[408]

The size of the inner castle of Dover appears to be about 6 acres,
reckoning the keep ward at 2, and the bailey at about 4.

The value of the town of Dover had trebled at the time of the Survey,
in spite of the burning of the town at William's first advent.[409]

[Illustration: FIG. 15. Dudley, Staffs. Dunster, Somerset.]

DUDLEY, Staffordshire (Fig. 15).--William Fitz Ansculf held Dudley
at the time of the Survey, "and there is his castle."[410] Mr Clark
appears to accept the dubious tradition of a Saxon Dodda, who first
built this castle in the 8th century, since he speaks of Dudley as
"a great English residence."[411] This tradition, however, is not
supported by Domesday Book, which shows Dudley to have been only a
small and unimportant manor before the Conquest. The strong position
of the hill was no doubt the reason why the Norman placed his castle
there. There is no Norman masonry in the present ruins. The earliest
work is that of the keep on the motte, a rectangular tower with round
corner turrets, attributed by Mr W. St John Hope to about 1320. The
first castle was demolished by Henry II. in 1175,[412] and an attempt
to restore it in 1218 was stringently countermanded.[413] The case
of Dudley is one of those which proves that Henry II. destroyed some
lawful castles in 1175 as well as the unlawful ones. In 1264 a license
to restore it was granted to Roger de Somery, in consideration of his
devotion to the king's cause in the Barons' War.[414] The whole area
of the castle, including the motte, but not including the works at
the base of the hill on which it stands, is 1-3/4 acres. The bailey
is an irregular oval, following the hill top. Dudley is an instance
in which the value of the manor has gone down instead of up since the
erection of the castle; this may perhaps be laid to the account of the
devastation caused through the Staffordshire insurrection of 1069.

DUNSTER, Somerset (Fig. 15).--Called Torre in Domesday Book. "There
William de Moion has his castle."[415] The motte here appears to be a
natural rock or _tor_, whose summit has been levelled and its sides
scarped by art. About 80 feet below the top is a (roughly) half-moon
bailey, itself a shelf on the side of the hill; there is another and
much smaller shelf at the opposite end.[416] Some foundations found in
the S.W. corner of the upper ward appear to indicate a former stone
keep.[417] Dunster was only a small manor of half a hide before the
Conquest, but afterwards its value tripled. There was a borough as
well as a castle.[418] The castle became the _caput baroniæ_ of the
De Moions, to whom the Conqueror gave fifty-six manors in different
parts of the county. There is not the slightest reason to suppose that
the site was fortified before the Conquest. Mr Clark remarks that "it
is remarkable that no mouldings or fragments of Norman ornament have
been dug up in or about the site, although there is original Norman
work in the parish church." The simple explanation, probably, is that
the first castle of De Moion was of wood, although on a site where it
would have been possible to build in stone from the first, as it does
not appear that any part of the motte is artificial. The area of the
bailey is 1-3/4 acres. The value of Dunster had risen at the date of
Domesday.[419]

[Illustration: FIG. 16. Durham.]

DURHAM (Fig. 16).--The castle here was first built by the Conqueror, on
his return from his expedition against Scotland in 1072.[420] It was
intended as a strong residence for the bishop, through whom William
hoped to govern this turbulent part of the country. He placed it on the
neck of the lofty peninsula on which the cathedral stands. The motte of
the Conqueror still remains, and so does the chapel[421] which he built
in the bailey; probably the present court of the castle, though crowded
now with buildings, represents the outline of the original bailey.[422]
The present shell keep on the motte was built by Bishop Hatfield in
Edward III.'s reign,[423] but has been extensively modernised. There
can be little doubt that up to 1345 there were only wooden buildings
on the motte, as the writer was informed by Canon Greenwell that no
remains of older stone-work than the 14th century had been found there.
It is so seldom that we get any contemporary description of a castle
of this kind, that it seems worth while to translate the bombastic
verse in which Laurence, Prior of Durham, described that of Durham in
Stephen's reign:[424]

"Not far hence [from the north road into the city] a tumulus of rising
earth explains the flatness of the excavated summit, explains the
narrow field on the flattened vertex, which the apex of the castle
occupies with very pleasing art. On this open space the castle is
seated like a queen; from its threatening height, it holds all that
it sees as its own. From its gate, the stubborn wall rises with the
rising mound,[425] and rising still further, makes towards the comfort
(amæna) of the keep. But the keep, compacted together, rises again
into thin air, strong within and without, well fitted for its work, for
within the ground rises higher by three cubits than without--ground
made sound by solid earth. Above this, a stalwart house[426] springs
yet higher than the [shell] keep, glittering with splendid beauty in
every part; _four posts are plain, on which it rests, one post at each
strong corner_.[427] Each face is girded by a beautiful gallery, which
is fixed into the warlike wall.[428] A bridge, rising from the chapel
[in the bailey] gives a ready ascent to the ramparts, easy to climb;
starting from them, a broad way makes the round of the top of the wall,
and this is the usual way to the top of the citadel.... The bridge
is divided into easy steps, no headlong drop, but an easy slope from
the top to the bottom. Near the [head of the] bridge, a wall descends
from the citadel, turning its face westward towards the river.[429]
From the river's lofty bank it turns away in a broad curve to meet the
field [_i.e._, Palace Green]. It is no bare plot empty of buildings
that this high wall surrounds with its sweep, but one containing
goodly habitations.[430] There you will find two vast palaces built
with porches, the skill of whose builders the building well reveals.
There, too, the chapel stands out beautifully raised on six pillars,
not over vast, but fair enough to view. Here chambers are joined to
chambers, house to house, each suited to the purpose that it serves....
There is a building in the middle of the castle which has a deep well
of abundant water.... The frowning gate faces the rainy south, a gate
that is strong, high-reaching, easily held by the hand of a weakling or
a woman. The bridge is let down for egress,[431] and thus the way goes
across the broad moat. It goes to the plain which is protected on all
sides by a wall, where the youth often held their joyous games. Thus
the castellan, and the castle artfully placed on the high ridge, defend
the northern side of the cathedral. And from this castle a strong wall
goes down southwards, continued to the end of the church."[432]

The original bailey of this castle covers 1 acre.

ELY, Cambridgeshire (Fig. 17).--This castle was built by William I. in
1070, when he was repressing the last struggle of the English under
the heroic Hereward. The monks of Ely felt it a sore grievance that he
placed the castle within their own bounds.[433] Both this castle and
the one built by William at Aldreth, to defend the passage into the
Isle of Ely, had a continuous existence, as they were both refortified
by Nigel, Bishop of Ely in Stephen's reign, and Ely Castle was besieged
and taken by Stephen.[434] The earthworks of this castle still exist,
to the south of the Minster. There is a fine motte with an oval
bailey, of which the banks and ditches are traceable in parts. The area
of the bailey is 2-1/2 acres. Of Aldreth or Aldrey there appear to be
no remains.

The value of the manor of Ely was £33 in the Confessor's reign; it fell
to £20 after the devastations of the Conquest, but had risen again to
£30 at the time of the Survey.[435]

[Illustration: FIG. 17. Ely, Cambs. Ewias Harold, Hereford. Eye,
Suffolk.]

EWIAS, Herefordshire (Fig. 17).--The brief notice of this castle
in Domesday Book throws some light on the general theory of
castle-building in England.[436] William FitzOsbern, as the king's
vicegerent, rebuilt this march castle, and committed it to the keeping
of another Norman noble, and the king confirmed the arrangement. But
in theory the castle would always be the king's. This is the only
case in the Survey where we hear of a castle being _rebuilt_ by the
Normans. We naturally look to one of King Edward's Norman favourites
as the first founder, for they alone are said by history to have built
castles on the Welsh marches before the Conquest. Dr Round conjectures
that Ewias was the "Pentecost's castle" spoken of in the (Peterborough)
_Anglo-Saxon Chronicle_ in 1052.[437] No masonry is now to be seen on
the motte at Ewias, but Mr Clark states that the outline of a circular
or polygonal shell keep is shown by a trench out of which the
foundations have been removed. The bailey is roughly of half-moon shape
and the mound oval. The whole area of the castle, including the motte
and banks, is 2-1/3 acres.

EXETER.--This castle is not mentioned in Domesday Book, but
Ordericus tells us that William _chose_ a site for the castle within
the walls, and left Baldwin de Molis, son of Count Gilbert, and
other distinguished knights, to finish the work, and remain as a
garrison.[438] In spite of this clear indication that the castle was
a new thing, it has been obstinately held that it only occupied the
site of some former castle, Roman or Saxon.[439] Exeter, of course,
was a Roman castrum, and its walls had been restored by Athelstan.
In this case William placed his castle inside instead of outside the
city walls, because, owing to the natural situation of Exeter, he
found in the north-west corner a site which commanded the whole city.
Although Domesday Book is silent about the castle, it tells us that
forty-eight houses in Exeter had been destroyed since William came to
England,[440] and Freeman remarks that "we may assume that these houses
were destroyed to make room for the castle, though it is not expressly
said that they were."[441]

Exeter Castle stands on a natural knoll, occupying the north-west
corner of the city, which has been converted into a sort of square
motte by digging a great ditch round the two sides of its base towards
the town.[442] That this ditch is no pre-Roman work is shown by the
fact that it stops short at the Roman wall, and begins again on the
outside of it, where, however, the greater part has been levelled to
form the promenade of the _Northernhay_ or north rampart of the city.
On top of this hill, banks 30 feet high were thrown up, which still
remain, and give to the courtyard which they enclose the appearance
of a pit.[443] On top of these banks there are now stone walls; but
these were certainly no part of the work of Baldwin de Molis, who must
have placed a wooden stockade on the banks which he constructed. One
piece of stonework he probably did set up, the gatehouse, which by its
triangle-headed windows and its long-and-short work is almost certainly
of the 11th century. It has frequently been called Saxon, but more
careful critics now regard it as "work that must have been done, if not
by Norman hands, at Norman bidding and on Norman design."[444] It was
no uncommon thing at this early period to have gatehouses of stone to
walls of earth and wood. Of these gatehouses Exeter is the most perfect
and the most clearly stamped with antiquity.

One thing we look for in vain at Exeter, and that is a citadel. There
is no keep, and there is no record that there ever was one, though a
chapel, hall, and other houses are mentioned in ancient accounts. Mr
Clark says that probably the Normans regarded the whole court as a
shell keep. It certainly was, in effect, a motte; but it was altogether
exceptional among Norman castles of importance if it had no bailey. And
in fact a bailey is mentioned in the _Pipe Roll_ of 1 Richard I., where
there is an entry for the cost of making a gaol in the bailey of the
castle.[445] Now Norden, who published a plan of Exeter in 1619, says
that the prison which formerly existed at the bottom of Castle Lane (on
the south or city front of the present castle) was "built upon Castle
grounde," and he states that the buildings and gardens which have been
made on this ground are intrusions on the king's rights.[446] The
remarkably full account of the siege of Exeter in the _Gesta Stephani_
speaks of an outer _promurale_ which was taken by Stephen, as well as
the inner bridge leading from the town to the castle, before the attack
on the castle itself. Unfortunately the word _promurale_ has the same
uncertainty about it that attaches to so many mediæval terms, and the
description given of it would apply either to the banks of a bailey,
or to the _heriçon_ on the counterscarp of the ditch of the motte.
We must, therefore, leave it to the reader's judgment whether the
evidence given above is sufficient to establish the former existence
of a bailey at Exeter, and to place Exeter among the castles of the
motte-and-bailey type.

The description of the castle given by the writer of the Gesta has
many points of interest.[447] He describes the castle as standing on a
very high mound (_editissimo aggere_) hedged in by an insurmountable
wall, which was defended by "Cæsarian" towers built with the very
hardest mortar. This must refer to Roman towers which may have existed
on the Roman part of the wall. Whether there was a stone wall on the
other two sides, facing the city, may be doubted, as the expenditure
entered to Henry II. in the _Pipe Rolls_ suggests that he was the
first to put stone walls on the banks, and the two ancient towers
which still exist appear to be of his time.[448] The chronicler goes
on to say that after Stephen had taken the _promurale_ and broken down
the bridge, there were several days and nights of fighting before he
could win the castle, which was eventually forced to surrender by the
drying-up of the wells. The mining operations which he describes were
no doubt undertaken with the view of shaking down the Roman wall at the
angle where it joins the artificial bank of Baldwin de Molis. Possibly
the chamber in the rock with the mysterious passages leading from it,
which is still to be seen in the garden of Miss Owthwaite, at the
point where the ditch ends, is the work of Stephen's miners.[449] The
description of his soldiers scrambling up the _agger_ on their hands
and knees (_quadrupede incessu_) will be well understood by those who
have seen the castle bank as it still rises from that ditch.

The present ward of Exeter Castle, which is rudely square in plan,
covers an area of 2 acres, which is as large as the whole area of many
of the smaller Norman castles. The castle was allowed to fall into
decay as early as 1549,[450] and since then it has been devastated by
the building of a Sessions House and a gaol. No plan has been preserved
of the former buildings in this court, though the site of the chapel is
known.

There is no statement in Domesday Book as to the value of Exeter.

EYE, Suffolk (Fig. 17).--This castle was built by William Malet, one
of the companions of the Conqueror, who is described as having been
half Norman and half English.[451] Eye, as its name implies, seems
to have been an island in a marsh in Norman times, and therefore a
naturally defensible situation. The references in the _Pipe Rolls_ to
the _palicium_ and the _bretasches_ of Eye Castle show that the outer
defences of the castle at any rate were of wood in the days of Henry
II.[452] That there were works in masonry at some subsequent period
is shown by a solitary vestige of a wing wall of flints which runs up
the motte. A modern tower now occupies the summit. The bailey of the
castle, the outline of which can still be traced, though the area is
covered with buildings and gardens, was oval in shape, and covered 2
acres.

The value of the manor of Eye had gone up since the Conquest from
£15 to £21. This must have been due to the castle and to the market
which Robert Malet or his son William established close to the castle;
for the stock on the manor and the number of ploughs had actually
decreased.[453] A proof that there is no deliberate register of
castles in Domesday Book is furnished by the very careful inventory
of the manor of Eye, where there is no mention of a castle, though it
is noticed that there are now a park and a market; and it is only in
the account of the lands of the bishop of Thetford, in mentioning the
injury which William Malet's market at Eye had done to the bishop's
market at Hoxne, that the castle of Eye is named.

GLOUCESTER.--"There were sixteen houses where the castle sits, but
now they are gone, and fourteen have been destroyed in the _burgus_
of the city," says Domesday Book.[454] Gloucester was undoubtedly a
Roman _chester_, and Roman pavements have been found there.[455] The
description in the Survey would lead us to think that the castle was
outside the ancient walls,[456] though Speed's map places it on the
line of the wall of his time, which may have been a mediæval extension.
The castle of Gloucester is now entirely destroyed, but there is
sufficient evidence to show that it was of the usual Norman type. There
was a motte, which was standing in 1819, and which was then called
the Barbican Hill;[457] it appears to have been utilised as part of
the works of the barbican. This motte must originally have supported
a wooden keep, and Henry I. must have been the builder of the stone
keep which Leland saw "in the middle of the area;"[458] for in 1100
Henry gave lands to Gloucester Abbey "in exchange for the site where
now the keep of Gloucester stands."[459] The bailey had previously
been enlarged by William Rufus.[460] Possibly the _framea turris_ or
framework tower spoken of in Henry II.'s reign may refer to the wooden
keep which had been left standing on the motte.[461] The walls of
Gloucester Castle were frequently repaired by Henry II.,[462] but the
word _murus_ by no means implies always a stone wall, and it is certain
that the castle was at that time surrounded by a wooden stockade, as
a writ of a much later period (1225) says that the stockade which is
around our castle of Gloucester has been blown down and broken by the
wind, and must be repaired.[463] Wooden bretasches on the walls are
spoken of in the _Pipe Rolls_ of 1193, and even as late as 1222.[464]

The value of the city of Gloucester had apparently risen at the time of
the Survey, though the entry being largely in kind, T. R. E., it is not
easy to calculate.

[Illustration: FIG. 18. Hastings, Sussex. Huntingdon.]

HASTINGS, Sussex (Fig. 18).--In this case we have positive contemporary
evidence that the earthen mound of the castle was thrown up by the
Normans at the time of the Conquest, for there is a picture in the
Bayeux Tapestry which shows them doing it. A number of men with
spades are at work raising a circular mound, on the top of which,
with the usual all-inclusiveness of mediæval picturing, a stockade
is already erected. A man with a pick seems to be working at the
ditch. The inscription attached is: "He commands that a castle be
dug at Hestengaceastra."[465] There is no need to comment on the
significance of this drawing and its inscription for the history of
early Norman castles; what is extraordinary is that it should have
been entirely overlooked for so long. In no case is our information
more complete than about Hastings. Not only does Domesday Book mention
the _castellaria_ of Hastings,[466] but the _Anglo-Saxon Chronicle_
also tells us that William built a castle there, while the chronicle
of Battle Abbey makes the evidence complete by telling us that "having
taken possession of a suitable site, he built a _wooden castle_
there."[467] This of course means the stockade on top of the motte,
with the wooden tower or towers which would certainly be added to it.
Wace states that this wooden castle was brought over in pieces in the
ships of the Count of Eu.[468]

The masonry now existing at the castle is probably none of it older
than the reign of Henry II. at the earliest, and most of it is
certainly much later.[469] The _Pipe Rolls_ show that Henry II. spent
£235 on the castle of Hastings between the years 1160 and 1181, and
it is indicated that some of this money was for stone, and some was
for a keep (_turrim_).[470] There is no tower large enough for a keep
at Hastings now, nor have any stone foundations been found on the
motte, and Mr Harold Sands, who has paid particular attention to this
castle, concludes that Henry II.'s keep has been carried away by the
sea, which has probably torn away at least 2 acres from the area of
the castle.[471] The beautiful fragment of the Chapel of St Mary is
probably of Henry II.'s reign; the walls and towers on the east side
of the castle appear to be of the 13th century. The ditch does not run
round the motte, but is cut through the peninsular rock on which the
castle stands, the motte and its ward being thus isolated. The form of
this bailey is now triangular, but it may have been square originally.
Beyond the ditch is another bailey, defended by earthen banks and by a
second ditch cut through the peninsula.[472] No exact estimate can be
given of the original area of the castle, as so much of the cliff has
been carried away by the sea.

Hastings itself had been a fortified town before the Norman Conquest,
and is one of those mentioned in the _Burghal Hidage_. The name
Hæstingaceaster, given to it in the _Anglo-Saxon Chronicle_ (1050), is
a proof that the Saxons used the name _chester_ for constructions of
their own, as no Roman remains have been found at Hastings. But the
Norman castle is outside the town, on a cliff which overlooks it. As in
the case of the other ports of Sussex, the castle was committed to an
important noble, in this case the Count of Eu.

The manor of Bexley, in which Hastings Castle stood, had been laid
waste at the Conquest; at the date of the Survey it was again rising in
value, though it had not reached the figure of King Edward's days.[473]

HEREFORD.--There can be little doubt that the castle of Hereford was
built by the Norman Ralph, Earl of Hereford, Edward the Confessor's
nephew, about the year 1048.[474] It was burnt by the Welsh in 1055,
after which Harold fortified the town with a dyke and ditch; but as Mr
Freeman remarks, it is not said that he restored the castle.[475] The
motte of Earl Ralph is now completely levelled, but it is mentioned
several times in documents of the 12th century,[476] and is described
in a survey of 1652, from which it appears that it had a stone keep
tower, as well as a stone breastwork enclosing a small ward.[477]
It stood outside the N.W. corner of the bailey, surrounded by its
own ditch; the site is still called Castle Hill. If the castle was
not restored before the Norman Conquest it was certainly restored
afterwards, as in 1067 we find the "men of the castle" fighting with
Edric Child and the Welsh. The castle appears to have had stone walls
by the time of Henry II., as the mention of a kiln for their repair
proves.[478] But these walls had wooden towers.[479] The timber ordered
in 1213 "ad hordiandum castellum nostrum de Hereford"[480] refers to
the wooden _alures_ or machicolations which were placed on the tops of
walls for the purpose of defending the bases.

Though Hereford was a private castle in the Confessor's reign, it was
claimed for the crown by Archbishop Hubert, the Justiciary, in 1197,
and continued to be a royal castle throughout the 13th century.[481]

The bailey of Hereford Castle still exists, with its fine banks; it is
kite-shaped and encloses 5-1/2 acres. The castle stood within the city
walls, in the south-east angle.

The value of Hereford appears to have greatly increased at the date of
the Survey.[482]

HUNTINGDON (Fig. 18).--"There were twenty houses on the site of the
castle, which are now gone."[483] Ordericus tells us that the castle
of Huntingdon was built by William on his return from his second visit
to York in 1068.[484] Huntingdon had been a walled town in Anglo-Saxon
times, and was very likely first fortified by the Danes, but was
repaired by Edward the Elder. As in the case of so many other towns,
the houses outside the walls had to pay geld along with those of the
city, and it was some of the former which were displaced by the new
Norman castle. Huntingdon was part of the patrimony of Earl Waltheof,
and came to the Norman, Simon de Senlis, through his marriage with
Waltheof's daughter and heiress. The line of Senlis ended in another
heiress, who married David, afterwards the famous king of Scotland;
David thus became Earl of Huntingdon. In the insurrection of the
younger Henry in 1174, William the Lion, grandson of David, took
sides with the young king, and consequently his castle was besieged
and taken by the forces of Henry II.,[485] and the king ordered it to
be destroyed. The _Pipe Rolls_ show that this order was carried out,
as they contain a bill for "hooks for pulling down the stockade of
Huntingdon Castle," and "for the work of the new castle at Huntingdon,
and for hiring carpenters, and crooks, and axes."[486] We learn from
these entries that the original castle of the Conquest had just been
replaced by a new one, very likely a new fortification of the old
mounds by William, in anticipation of the insurrection. We also learn
that the new castle was a wooden one; for a castle which has to be
pulled down by carpenters with hooks and axes is certainly not of
stone. It does not appear that the castle was ever restored, though
"the chapel of the castle" is spoken of as late as the reign of Henry
III.[487]

The motte of Huntingdon still exists, and has not the slightest sign of
masonry. The bailey is roughly square, with the usual rounded corners;
the motte was inside this enclosure, but had its own ditch. The whole
area was 2-1/2 acres, but another bailey was subsequently added.

The value of Huntingdon appears to have been stationary at the time
of the Survey, the loss of the twenty houses causing a diminution of
revenue which must have been made up from the new feudal dues of the
castle.

LAUNCESTON, or Dunheved,[488] Cornwall (Fig. 19).--There, says Domesday
Book, is the castle of the Earl of Mortain.[489] In another place it
tells us that the earl gave two manors to the bishop of Exeter "for
the exchange of the castle of Cornwall," another name for Dunheved
Castle. We have already had occasion to note that the "exchange of the
castle," in Domesday language, is an abbreviation for the exchange of
the site of the castle. The fact that the land was obtained from the
church is a proof that the castle was new, for it was not the custom of
Saxon prelates thus to fortify themselves. The motte of Launceston is
a knoll of natural rock, which has been scarped and heightened by art.
This motte now carries a circular keep, which cannot be earlier than
the 13th century.[490] There is no early Norman work whatever about
the masonry of the castle, and the remarkably elaborate fortifications
on the motte belong to a much later period.[491] The motte rises in
one corner of a roughly rectangular bailey, which covers 3 acres. It
stands outside the town walls, which still exist, and join those of
the castle, as at Totnes. Launceston was only a small manor of ten
ploughs in the time of the Confessor. In spite of the building of
the castle, the value of the manor had greatly gone down in William's
time.[492] The ten ploughs had been reduced to five.

[Illustration: FIG. 19. Launceston, Cornwall. Lewes, Sussex.]

LEWES, Sussex (Fig. 19).--The castle of Lewes is not mentioned in its
proper place in Sussex by Domesday Book, and this is another proof that
the Survey contains no inventory of castles; for that the castle was
existing at that date is rendered certain by the numerous allusions
in the Norfolk portion to "the exchange of the castle of Lewes."[493]
It is clear that at some period, possibly during the revolt of Robert
Curthose in 1079, William I. gave large estates in Norfolk to his
trusty servant, William de Warenne, in exchange for the important
castle of Lewes, which he may have preferred to keep in his own hands
at that critical period. This bargain cannot have held long, at least
as regards the castle, which continued to belong to the Warenne family
for many generations. We cannot even guess now how the matter was
settled, but the lands in Norfolk certainly remained in the hands of
the Warennes.

Lewes is one of the very few castles in England which have two
mottes.[494] They were placed at each end of an oval bailey, each
surrounded by its own ditch, and each projecting about three-fourths
beyond the line of the bailey. On the northern motte only the
foundations of a wall round the top remain; on the other, part of the
wall which enclosed a small ward, and two mural towers. These towers
have signs of the early Perpendicular period, and are very likely of
the reign of Edward III., when the castle passed into the hands of
the Fitz Alans. The bailey, which enclosed an area of about 3 acres,
is now covered with houses and gardens, but parts of the curtain wall
on the S.E. and E. stand on banks, bearing witness to the original
wooden fortifications. The great interest of this bailey is its ancient
Norman gateway. The entrance was regarded by mediæval architects as the
weakest part of the fortress, and we frequently find that it was the
first part to receive stone defences.[495] It is not surprising that
at such an important place as Lewes, which was then a port leading to
Normandy, and at the castle of so powerful a noble, we should find an
early case of stone architecture supplementing the wooden defences.
But the two artificial mottes have no masonry that can be called early
Norman.

Lewes is one of the boroughs mentioned in the _Burghal Hidage_, and was
a _burgus_ at the time of the Survey.[496] The value of the town had
increased by £1, 18s. from what it had been in King Edward's time.

[Illustration: FIG. 20. LINCOLN.]

LINCOLN (Fig. 20).--Domesday Book tells us that 166 houses were
destroyed to furnish the site of the castle.[497] The _Anglo-Saxon
Chronicle_ says that William built a castle here on his return from
his first visit to York in 1068, and Ordericus makes the same
statement.[498] Lincoln, like Exeter, was a Roman _castrum_, and the
Norman castle in both cases was placed in one corner of the castrum;
but the old Roman wall of Lincoln, which stands on the natural ground,
was not considered to be a sufficient defence on the two exterior
sides, probably on account of its ruinous condition. It was therefore
buried in a very high and steep bank, which was carried all round the
new castle.[499] This circumstance seems to point to the haste with
which the castle was built, Lincoln being then for the first time
subdued. The fact that it was inside the probably closely packed Roman
walls explains why so many houses were destroyed for the castle.[500]
Lincoln, like Lewes, has two mottes: both are of about the same height,
but the one in the middle of the southern line of defence is the larger
and more important; it was originally surrounded with its own ditch. It
is now crowned with a polygonal shell wall, which may have been built
by the mother of Ralph Gernon, Earl of Chester, in the reign of Henry
I.[501] The tower on the other motte, at the south-east corner, has
been largely rebuilt in the 14th century and added to in modern times,
but its lower storey still retains work of Norman character. There is
good reason to suppose that this bailey was first walled with stone
in Richard I.'s reign, as there is an entry in the _Pipe Rolls_ of
1193-1194 "for the cost of fortifying the bailey, £82, 16s. 4d."[502]
The present wall contains a good deal of herring-bone work, and this
circumstance led Mr Clark, who was looking for something which he
_could_ put down to William I.'s time, to believe that the walls were
of that date. But the herring-bone work is all in patches, as though
for repairs, and herring-bone work was used for repairs at all epochs
of mediæval building. The two gateways (that is the Norman portions of
them) are probably of about the same date as the castle wall. The whole
area is 5-3/4 acres.

The total revenue which the city of Lincoln paid to the king and the
earl had gone up from 30_l._ T. R. E. to 100_l._ T. R. W. For the sake
of those who imagine that Saxon halls had anything to do with mottes,
it is worth noting that the hall which was the residence of the chief
landholder in Lincoln before the Conquest was still in existence after
the building of the castle, but evidently had no connection with
it.[503]

[Illustration: FIG. 21. Monmouth. Montacute, Somerset. Morpeth,
Northumberland.]

MONMOUTH (Fig. 21).--Domesday Book says that the king has four ploughs
in demesne in the castle of Monmouth.[504] Dr Round regards this as
one of the cases where _castellum_ is to be interpreted as a town and
not as a castle. However this may be, the existence of a Norman castle
at Monmouth is rendered certain by a passage in the _Book of Llandaff_,
in which it is said that this castle was built by William FitzOsbern,
and a short history of it is given, which brings it up to the days of
William Fitz Baderun.[505] Speed speaks of this castle as "standing
mounted round in compasse, and within her walls another mount, whereon
a Towre of great height and strength is built."[506] This sounds like
the description of a motte and bailey; but the motte cannot be traced
now. It is possible that it may have been swept away to build the
present barracks; the whole castle is now on a flat-topped hill. The
area is 1-3/4 acres.[507]

The value of the manor before the Conquest is not given.

MONTACUTE, Somerset (Fig. 21).--This is another instance of a site for
a castle obtained by exchange from the church. Count Robert of Mortain
gave the manor of Candel to the priory of Athelney in exchange for
the manor of Bishopstowe, "and there is his castle, which is called
Montagud."[508] The English name for the village at the foot of the
hill was Ludgarsburh, which does not point to any fortification on the
hill itself, the spot where the wonder-working crucifix of Waltham was
found in Saxon times. Robert of Mortain's son William gave the castle
of Montacute, with its chapel, orchard, and other appurtenances, to a
priory of Cluniac monks which he founded close to it. The gift may have
had something compulsory in it, for William of Mortain was banished
by Henry I. in 1104 as a partisan of Robert Curthose. Thus, as Leland
says, "the notable castle partly fell to ruin, and partly was taken
down to make the priory, so that many years since no building of it
remained; only a chapel was set upon the very top of the dungeon, and
that yet standeth there."[509] There is still a high oval motte, having
a ditch between its base and the bailey; the latter is semilunar in
shape. The hill has been much terraced on the eastern side, but this
may have been the work of the monks, for purposes of cultivation.[510]
There is no masonry except a quite modern tower. According to Mr Clark,
the motte is of natural rock. The French name of the castle was of
course imported from Normandy, and we generally find that an English
castle with a Norman-French name of this kind has a motte.[511]

Bishopstowe, in which the castle was placed, was not a large manor in
Saxon times. Its value T. R. E. is not given in the Survey, but we
are told that it is worth 6_l._ to the earl, and 3_l._ 3_s._ to the
knights who hold under him.

MORPETH, Northumberland (Fig. 21).--There is only one mention known
to us of Morpeth Castle in the 11th century, and that is in the
poem of Geoffrey Gaimar.[512] He says that William Rufus, when
marching to Bamborough, to repress the rebellion of Mowbray, Earl of
Northumberland, "took the strong castle of Morpeth, which was seated
on a little mount," and belonged to William de Morlei. Thus there
can be no doubt that the Ha' Hill, about 100 yards to the N. of the
present castle, was the motte of the first castle of Morpeth, though
the remains of the motte, which are mentioned by Hodgson, have been
destroyed.[513] A natural ridge has been used to form a castle by
cutting off its higher end to form a motte, and making a court on the
lower part of the ridge. The great steepness of the slopes rendered
ordinary ditches unnecessary, nor are there any traces now of banks or
foundations. In the court some Norman capitals and carved stones were
found in 1830. This early castle was admirably placed for commanding
the river and the bridge.[514] The present castle of Morpeth was built
in 1342-1349.[515]

NEWCASTLE, Northumberland.--The first castle here was built by
Robert, son of William I., on his return from his expedition to
Scotland in 1080.[516] It was of the usual motte-and-bailey kind, the
motte standing in a small bailey which was rectilinear and roughly
oblong.[517] This motte was in existence when Brand wrote his _History
of Newcastle_, but was removed in 1811. The castle was placed outside
the Roman station at Monkchester, and commanded a Roman bridge over the
Tyne, "and to the north-east overlooked a ravine that under the name of
The Side formed for centuries a main artery of communication between
England and Scotland."[518] Henry II., when he built the fine keep of
this castle, did not place it on the motte, but in the outer and larger
ward, which was roughly triangular. The outer curtain appears to have
stood on the banks of the former earthen castle, as the Parliamentary
Survey of 1649 speaks of the castle as "bounded with strong works of
stone and mud."[519] The area of the whole castle was 3 acres and 1
rood.

[Illustration: FIG. 22. Norham. Nottingham.]

NORHAM, Northumberland (Fig. 22).--The first castle here was built
by Ranulf Flambard, Bishop of Durham, in the reign of William Rufus.
It was built to defend Northumberland against the incursions of the
Scots, and we are expressly told that no castle had existed there
previously.[520] This first castle, which we may certainly assume to
have been of earth and wood, was destroyed by the Scots in 1138, and
there does not seem to have been any stone castle until the time of
Bishop Puiset or Pudsey, who built the present keep by command of King
Henry II.[521] Mr Clark tried hard to find some work of Flambard's in
this tower, but found it difficult, and was driven back on the rather
lame assumption that "the lapse of forty [really fifty at least] years
had not materially changed the style of architecture then in use."[522]
In fact, the Norman parts of this keep show no work so early as the
11th century, but are advanced in style, for not only was the basement
vaulted, but the first floor also. The simple explanation is that
Flambard threw up the large square motte on which the keep now stands,
and provided it with the usual wooden defences. It also had a strong
tower, but almost certainly a wooden one; hence it was easily destroyed
by the Scots when once taken.[523] The motte was probably lowered to
some extent when the stone keep was built. It stands on a high bank
overlooking the Tweed, and is separated from its bailey by a deep
ditch. The bailey may be described as a segment of a circle; its area
is about 2 acres.

NORWICH (Fig. 23).--We find from Domesday Book that no less than 113
houses were destroyed for the site of this castle, a certain proof that
the castle was new.[524] It is highly probable that it was outside the
primitive defences of the town, at any rate in part. Norwich was built,
partly on a peninsula formed by a double bend of the river Wensum,
partly in a district lying south-west of this peninsula, and defended
by a ridge of rising ground running in a north-easterly direction. The
castle was placed on the edge of this ridge, and all the oldest part
of the town, including the most ancient churches, lies to the east of
it.[525] In the conjectural map of Norwich in 1100, given in Woodward's
_History of Norwich Castle_,[526] the street called Burg Street divides
the Old Burg on the east from the New Burg on the west; this street
runs along a ridge which traverses the neck of the peninsula from
south-west to north-east, and on the northern end of this ridge the
castle stands.[527] There can be little doubt that this street marks
the line of the _burh_ or enclosing bank by which the primitive town
of Norwich was defended.[528] A clear proof of this lies in the fact
that the castle of Norwich was anciently not in the jurisdiction of the
city, but in that of the county; the citizens had no authority over the
houses lying beyond the castle ditches until it was expressly granted
to them by Edward III.[529] The mediæval walls of Norwich, vastly
extending the borders of the city, were not built till Henry III.'s
reign.[530]

[Illustration: FIG. 23. Norwich.

(From Harrod's "Gleanings among the Castles and Convents of Norfolk,"
p. 133.)]

The motte of Norwich Castle, according to recent investigations,
is entirely artificial;[531] it was originally square, and had "a
prodigious large and deep ditch around it."[532] The fancy of the
antiquary Wilkins that the motte was the centre of two concentric
outworks[533] was completely disproved by Mr Harrod, who showed that
the original castle was a motte with one of the ordinary half-moon
baileys attached. Another ward, called the Castle Meadow, was probably
added at a later date. The magnificent keep which now stands on the
motte is undoubtedly a work of the 12th century.[534] The castle which
Emma, wife of Earl Ralf Guader, defended against the Conqueror after
the celebrated bride-ale of Norwich was almost certainly a wooden
structure. As late as the year 1172 the bailey was still defended by
a wooden stockade and wooden bretasches;[535] and even in 1225 the
stockade had not been replaced by a stone wall.[536]

Norwich was a royal castle, and consequently always in the hands of the
sheriff; it was never the property of the Bigods.[537] As the fable
that extensive lands belonging to the monastery of Ely were held on the
tenure of castle guard at Norwich _before the Conquest_ is repeated
by all the local historians,[538] it is worth while to note that the
charters of Henry I. setting the convent free from this service, make
no allusion to any such ancient date for it,[539] and that the tenure
of castle guard is completely unknown to the Anglo-Saxon laws. The area
of the inner bailey is 3-1/4 acres, and that of the outer, 4-1/2 acres.
The value of Norwich had greatly risen since the Conquest.[540]

NOTTINGHAM (Fig. 22).--This important castle is not mentioned in
Domesday Book, but the _Anglo-Saxon Chronicle_ says that William I.
built the castle at Nottingham in 1067, on his way to repress the
first insurrection in Yorkshire. Ordericus, repeating this statement,
adds that he committed it to the keeping of William Peverel.[541]
The castle was placed on a lofty headland at some distance from the
Danish borough, and between the two arose the Norman borough which is
mentioned in Domesday Book as the _novus burgus_. The two upper wards
of the present castle probably represent William's plan. The upper ward
forms a natural motte of rock, as it is 15 feet higher than the bailey
attached to it, and has been separated from it by a ditch cut across
the rocky headland, which can still be traced below the modern house
which now stands on the motte. Such a site was not only treated as a
motte, but was actually called by that name, as we read of the _mota_
of Nottingham Castle in the _Pipe Rolls_ of both John's and Richard
I.'s reigns.

Mr Clark published a bird's-eye view of Nottingham Castle in his
_Mediæval Military Architecture_, about which he only stated that it
was taken from the _Illustrated London News_. It does not agree with
the plan made by Simpson in 1617,[542] and is therefore not quite
trustworthy; the position of the keep, for example, is quite different.
The keep, which Hutchison in his Memoirs speaks of as "the strong tower
called the Old Tower on the top of the rock," seems clearly Norman,
from the buttresses. It was placed (according to Simpson's plan), on
the north side of the small ward which formed the top of the motte,
and was enclosed in a yet older shell wall which has now disappeared.
The height of this motte is indicated in the bird's-eye view by the
ascending wall which leads up it from the bailey. It had its own ditch,
as appears by several mentions in the accounts of "the drawbridge of
the keep," and "the bridge leading up to the dongeon."[543] It is
highly probable that this keep was built by King John, as in a _Mise
Roll_ of 1212 there is a payment entered "towards making the tower
which the king commanded to be built on the motte of Nottingham."[544]
But the first masonry in the castle was probably the work of Henry
II., who spent £1737, 9s. 5d. on the castle and houses, the gaol, the
king's chamber, the hall, and in raising the walls and enclosing the
bailey.[545] The castle has been so devastated by the 17th century
spoiler, that the work of Henry and John has been almost entirely
swept away, but the one round tower which still remains as part of
the defences of the inner bailey, looks as though it might be of the
time of Henry II. This bailey is semicircular; the whole original
castle covers only 1-2/3 acres. A very much larger bailey was added
afterwards, probably in John's reign.[546] Probably this later bailey
was at first enclosed with a bank and stockade, and this stockade may
be the palitium of which there are notices in the records of Henry
III. and Edward I.[547] The main gateway of this bailey, which still
remains, is probably of Edward I. or Edward II.'s reign.[548]

The castle of Nottingham was the most important one in the Midlands,
and William of Newburgh speaks of it as "so well defended by nature and
art that it appears impregnable."[549] The value of the town had risen
from £18 to £30 at the time of the Survey.[550]

[Illustration: FIG. 24. Okehampton, Devon. Penwortham, Lancs. Pevensey,
Sussex.]

OKEHAMPTON, Devon (Fig. 24).--Baldwin de Molis, Sheriff of Devon, held
the manor of Okehampton at the time of the Survey, and had a castle
there.[551] On a hill in the valley of the Okement River stand the
remains of a castle of the motte-and-bailey pattern. On the motte,
which is high and steep, are the ruins of a keep of late character,
probably of the 14th century.[552] The oval bailey covers 1/2 an acre,
and the whole castle is surrounded with a very deep ditch (filled up
now on the east side) which is in part a natural ravine. The usual
ditch between the motte and the bailey is absent here. This castle
appears to have continued always in private hands, and therefore there
is little to be learned about it from the public records. The value of
Okehampton manor had increased since the Conquest from £8 to £10. As
there is no _burgus_ mentioned T. R. E., but four _burgenses_ and a
market T. R. W., Baldwin the Sheriff must have built a borough as well
as a castle. Otherwise it was a small manor of thirty ploughs.

OSWESTRY, Shropshire.--Mr Eyton's identification of the Domesday castle
of Louvre, in the manor of Meresberie, Shropshire, with Oswestry, seems
to be decisive.[553] The name is simply L'Oeuvre, the Work, a name
very frequently given to castles in the early Norman period. Domesday
Book says that Rainald de Bailleul built a castle at this place.[554]
He had married the widow of Warin, Sheriff of Shropshire, who died in
1085. The castle afterwards passed into the hands of the Fitz Alans,
great lords-marcher on the Welsh border. As the Welsh annals give the
credit of building the castle to Madoc ap Meredith, into whose hands it
fell during the reign of Stephen, it is not impossible that some of the
masonry still existing on the motte, which consists of large cobbles
bedded in very thick mortar, may be his work, and probably the first
stonework in the castle. A sketch made in the 18th century, however,
which is the only drawing preserved of the castle, seems to show
architecture of the Perpendicular period.[555] But probably the keep
alone was of masonry in the 12th century, as in 1166, when the castle
was in royal custody, the repair of the stockade is referred to in the
_Pipe Rolls_.[556] No plan has been preserved of Oswestry Castle, so
that it is impossible to recover the shape or area of the bailey, which
is now built over. The manor of Meresberie had been unoccupied (wasta)
in the days of King Edward, but it yielded 40s. at the date of the
Survey. Eyton gives reasons for thinking that the town of Oswestry was
founded by the Normans.

[Illustration: FIG. 25. Oxford.

(From "Oxonia Illustrata," David Loggan, 1675.)]

OXFORD (Fig. 25).--This castle was built in 1071 by Robert d'Oilgi (or
d'Oilly), a Norman who received large estates in Oxfordshire.[557]
Oxford was a burgus in Saxon times, and is one of those mentioned in
the _Burghal Hidage_. Domesday tells us that the king has twenty mural
mansions there, which had belonged to Algar, Earl of Mercia, and that
they were called mural mansions because their owners had to repair the
city wall at the king's behest, a regulation probably as old as the
days of Alfred. The Norman castle was placed outside the town walls,
but near the river, from which its trenches were fed.[558] It was
without doubt a motte-and-bailey castle; the motte still remains, and
the accompanying bird's-eye view by David Loggan, 1675, shows that the
later stone walls of the bailey stood on the earthen banks of D'Oilly's
castle. The site is now occupied by a gaol. On the line of the walls
rises the ancient tower of St George's Church, which so much resembles
an early Norman keep that we might think it was intended for one, if
the Osney chronicler had not expressly told us that the church was
founded two years after the castle.[559] It is evident that the design
was to make the church tower work as a mural tower, a combination of
piety and worldly wisdom quite in accord with what the chronicler tells
us of the character of Roger d'Oilly.

Henry II. spent some £260 on this castle between the years 1165 and
1173, the houses in the keep, and the well being specially mentioned.
We may presume that he built with stone the decagonal [shell?] keep on
the motte, whose foundations were discovered at the end of the 18th
century.[560] There is still in the heart of the motte a well in a very
remarkable well chamber, the masonry of which may be of his time. The
area of the bailey appears to have been 3 acres.

The value of the city of Oxford had trebled at the time of the Domesday
Survey.[561]

In the treaty between Stephen and Henry in 1153 the whole castle of
Oxford is spoken of as the "Mota" of Oxford.[562]

PEAK CASTLE, Derbyshire.--The Survey simply calls this castle the
Castle of William Peverel, but tells us that two Saxons had formerly
held the _land_.[563] There is no motte here, but the strong position,
defended on two sides by frightful precipices, rendered very little
fortification necessary. It is possible that the wall on the N. and
W. sides of the area may be, in part at least, the work of William
Peverel; the W. wall contains a great deal of herring-bone work, and
the tower at the N.W. angle does not flank at all, while the other
one in the N. wall only projects a few feet; the poor remains of
the gatehouse also appear to be Norman. It would probably be easier
to build a wall than to raise an earthbank in this stony country;
nevertheless, behind the modern wall which runs up from the gatehouse
to the keep, something like an earthbank may be observed on the edge of
the precipice, which ought to be examined before any conclusions are
determined as to the first fortifications of this castle. The keep,
which is of different stone to the other towers and the walls, stands
on the highest ground in the area, apparently on the natural rock,
which crops up in the basement. It is undoubtedly the work of Henry
II., as the accounts for it remain in the _Pipe Rolls_, and the slight
indications of style which it displays, such as the nook-shafts at the
angles, correspond to the Transition Norman period.[564] The shape of
the bailey is a quadrant; its area scarcely exceeds 1 acre.

The value of the manor had risen since the Conquest, and William
Peverel had doubled the number of ploughs in the demesne. The castle
only remained in the hands of the Peverels for two generations, and was
then forfeited to the crown. The manor was only a small one; and the
site of the castle was probably chosen for its natural advantages and
for the facility of hunting in the Peak Forest.

PENWORTHAM, Lancashire (Fig. 24).--"King Edward held Peneverdant. There
are two carucates of land there, and they used to pay ten pence. Now
there is a castle there, and there are two ploughs in the demesne,
and six burghers, and three radmen, and eight villeins, and four
cowherds. Amongst them all they have four ploughs. There is half a
fishery there. There is wood and hawk's eyries, as in King Edward's
time. It is worth £3."[565] The very great rise in value in this
manor shows that some great change had taken place since the Norman
Conquest. This change was the building of a castle. The _modo_ of
Domesday always expresses a contrast with King Edward's time, and
clearly tells us here that Penwortham Castle was new.[566] It lay in
the extensive lands between the Ribble and the Mersey, which were part
of the Conqueror's enfeoffment of Roger the Poitevin, third son of Earl
Roger de Montgomeri.[567] Since Penwortham is mentioned as demesne,
and no under-tenant is spoken of, we may perhaps assume that this
castle, which was the head of a barony, was built by Roger himself.
He did not hold it long, as he forfeited all his estates in 1102. At
a later period, though we have not been able to trace when, the manor
of Penwortham passed into the hands of the monks of Evesham, to whom
the church had already been granted, at the end of the Conqueror's
reign.[568] Probably it is because the castle thus passed into the
hands of the church that it never developed into a stone castle, like
Clitheroe. The seat of the barony was transferred elsewhere, and
probably the timbers of the castle were used in the monastic buildings
of Penwortham Priory.

The excavations which were made here in 1856 proved conclusively that
there were no stone foundations on the Castle Hill at Penwortham.[569]
These excavations revealed the singular fact that the Norman had thrown
up his motte on the site of a British or Romano-British hut, without
even being aware of it, since the ruins of the hut were buried 5 feet
deep and covered by a grass-grown surface, on which the Norman had laid
a rude pavement of boulders before piling his motte.[570] Among the
objects found in the excavations was a Norman prick spur, a conclusive
proof of the Norman origin of the motte.[571] No remains appear to have
been found of the Norman wooden keep; but this would be accounted for
by the theory suggested above.

Penwortham is a double motte, the artificial hill rising on the back of
a natural hill, which has been isolated from its continuing ridge by an
artificial ditch cut through it. The double hill rises out of a bailey
court which is rudely square, but whose shape is determined by the
ground, which forms a headland running out into the Ribble. The whole
area cannot certainly be ascertained. There was a ferry at this point
in Norman times.[572] The castle defends the mouth of the Ribble and
overlooks the town of Preston.

Penwortham was certainly not the _caput_ of a large soke in Saxon
times, as it was only a berewick of Blackburn, in which hundred it lay.
It was the Norman who first made it the seat of a barony.

PETERBOROUGH.--The chronicler, Hugh Candidus, tells us that Abbot
Thorold, the Norman abbot whom William I. appointed to the ancient
minster of Peterborough, built a castle close to the church, "which in
these days is called Mount Torold."[573] This mount is still existing,
but it has lost its ancient name, and is now called Tout Hill. It
stands in the Deanery garden, and has probably been largely ransacked
for garden soil, as it has a decayed and shapeless look. Still, it is a
venerable relic of Norman aggression, well authenticated.

PEVENSEY, Sussex (Fig. 24).--The Roman castrum of Pevensey (still
so striking in its remains) was an inhabited town at the date of
the Norman Conquest, and was an important port.[574] After taking
possession of the castrum, William I. drew a strong bank across its
eastern end, and placed a castle in the area thus isolated. This first
castle was probably entirely of wood, as there was a wooden _palicium_
on the bank as late as the reign of Henry II.[575] But if a wooden keep
was built at first, it was very soon superseded by one of stone.[576]
The remains of this keep have recently been excavated by Mr Harold
Sands and Mr Montgomerie, and show it to have been a most remarkable
building[577] (see Chapter XII., p. 355)--in all probability one
of the few 11th century keeps in England. We may perhaps attribute
this distinction to the fact that no less a man than the Conqueror's
half-brother, the Count of Mortain, was made the guardian of this
important port.

Pevensey is mentioned as a port in the _Close Rolls_ of Henry III.'s
reign, and was one of the important waterways to the Continent.[578] As
has been already noted, the establishment of the castle was followed
by the usual rise in the value of the _burgus_.[579] The area of the
castle covers 1 acre.

PONTEFRACT, Yorkshire (Fig. 26).--This castle is not spoken of in
Domesday by its French name, but there can be no doubt that it is "the
Castle of Ilbert" which is twice mentioned and several times alluded to
in the _Clamores_, or disputed claims, which are enrolled at the end of
the list of lands in Yorkshire belonging to the tenants-in-chief.[580]
The existence of Ilbert's castle at Pontefract in the 11th century is
made certain by a charter (only an early copy of which is now extant)
in the archives of the Duchy of Lancaster, in which William Rufus at
his accession regrants to Ilbert de Lacy "the custom of the castelry
of his castle, as he had it in the Conqueror's days and in those of
the bishop of Bayeux."[581] As Mr Holmes remarks, this carries us back
to four years before the compilation of Domesday Book, since Odo,
Bishop of Bayeux, whom William had left as regent during his absence in
Normandy, was arrested and imprisoned in 1082.[582]

Pontefract is called Kirkby in some of the earlier charters, and this
was evidently the English (or rather the Danish) name of the place. It
lay within the manor of Tateshall, which is supposed to be the same
as Tanshelf, a name still preserved in the neighbourhood of, but not
exactly at, Pontefract.[583] Tanshelf claims to be the Taddenescylf
mentioned in the _Anglo-Saxon Chronicle_, where King Edgar received
the submission of the Yorkshire Danes in 947. There is no proof that
the hill at Kirkby was fortified before the Conquest. It was a steep
headland rising out of the plain of the Aire, and needing only to be
scarped by art and to have a ditch cut across its neck to be almost
impregnable. It lay scarcely a mile east of the Roman road from
Doncaster to Castleford and the north.

[Illustration: FIG. 26. Pontefract, Yorks. Preston Capes, Northants.
Quatford, Salop.]

It is no part of our task to trace the fortunes of this famous
castle, which was considered in the Middle Ages to be the key of
Yorkshire.[584] In spite of the labels affixed to the walls we venture
to assert with confidence that none of the masonry now visible
belongs to the days of Ilbert. The structural history of the castle
was probably this: Ilbert de Lacy, one of the greatest of the Norman
tenants-in-chief in Yorkshire,[585] built in this naturally defensive
situation a castle of earth and wood, like other Norman castles.
Whether he found the place already defended by earthen banks we do not
attempt to decide, but analogy makes it fairly certain that the motte
was his work, and was crowned by a wooden tower. This motte, which
was at least partially scarped out of the soft sandstone rock, is now
disguised by the remarkable keep which has been built up around it,
consisting at present of two enormous round towers and the ruins of a
third. As a fourth side is vacant, it may reasonably be conjectured
that there was a fourth roundel.[586] If the plan was a quatrefoil
it resembled that of the keep of York, which is now ascertained to
belong to the reign of Henry III.; and the very little detail that is
left supports the view that Pontefract keep was copied from the royal
experiment at York, though it differed from it in that it actually
revetted the motte itself. There is no ditch now round the motte, but
we venture to think that its inner ditch is indicated by the position
of the postern in Piper's Tower, which seems to mark its outlet.
It appears to have been partly filled up during the great siege of
Pontefract in 1648.[587] The platform which is attached to the motte
on the side facing the bailey is probably an addition of the same
date, intended for artillery; its retaining wall shows signs of hasty
construction. A well chamber and a passage leading both to it and to a
postern opening towards the outer ditch appear to have been made in the
rocky base of the motte in the 13th century.

The area of the inner and probably original bailey of this castle,
including the motte, is 2-1/3 acres. The Main Guard, and another bailey
covering the approach on the S. side, were probably later additions,
bringing up the castle area to 7 acres. The shape of the first bailey
is an irregular oval, determined by the hill on which it stands.

The value of the manor of Tateshall had fallen at the time of the
Survey from £20 to £15, an unusual circumstance in the case of a manor
which had become the seat of an important castle; but the number of
ploughs had decreased by half, and we may infer that Tateshall had not
recovered from the great devastation of Yorkshire in 1068.[588]

PRESTON CAPES, Northants (Fig. 26).--That a castle of the 11th century
stood here is only proved by a casual mention in the _Historia
Fundationis_ of the Cluniac priory of Daventry, which tells us
that this priory was first founded by Hugh de Leycestre, Seneschal
of Matilda de Senlis, close to his own castle of Preston Capes,
about 1090. Want of water and the proximity of the castle proving
inconvenient, the priory was removed to Daventry.[589] The work lies
about 3 miles from the Watling Street. The castle stands on a spur of
high land projecting northwards towards a feeder of the river Nesse,
about 3 miles W. of the Watling Street. The works consist of a motte,
having a flat top 80 to 90 feet in diameter, and remains of a slight
breastwork. This motte is placed on the edge of the plateau, and the
ground falls steeply round its northern half. About 16 feet down this
slope, a ditch with an outer bank has been dug, embracing half the
mound. Lower down, near the foot of the slope, is another and longer
ditch and rampart. It is probable that the bailey occupied the flatter
ground S.E. of the motte, but the site is occupied by a farm, and no
traces are visible.[590]

The value of the manor of Preston Capes had risen from 6s. to 40s.
at the time of the Survey. It was held by Nigel of the Count of
Mellent.[591]

QUATFORD, Shropshire (Fig. 26).--There can hardly be any doubt that the
_nova domus_ at Quatford mentioned in the Survey was the new castle
built by Roger de Montgomeri, Earl of Shrewsbury. We have already
suggested that the _burgus_ which also existed there may have been his
work, and not that of the Danes.[592] The manor belonged to the church
before the Conquest.[593] The oval motte, which still remains, is
described as placed on a bold rocky promontory jutting into the Severn;
it is not quite 30 feet high, and about 60 feet by 120 in diameter
on top, and has a small bean-shaped bailey of 1 acre. It is near the
church, which has Norman remains.[594] Robert Belesme, son of Earl
Roger, removed the castle to Bridgenorth, and so the Quatford castle is
heard of no more.[595] The manor of Quatford was paying nothing at the
date of the Survey.

RAYLEIGH, Essex (Fig. 27).--"In this manor Sweyn has made his
castle."[596] Sweyn was the son of Robert Fitz-Wymarc, a half English,
half Norman favourite of Edward the Confessor. Robert was Sheriff of
Essex under Edward and William, and Sweyn appears to have succeeded
his father in this office.[597] Sweyn built his castle on land which
had not belonged to his father, so Rayleigh cannot be the "Robert's
Castle" of the _Anglo-Saxon Chronicle_, to which some of the Norman
adventurers fled on the triumph of Earl Godwin.[598] There is a fine
motte at Rayleigh, and a semicircular bailey attached; the ditch round
the whole is still well marked. There is not a vestige of masonry
on the surface, but some excavations made in 1910 revealed stone
foundations. The inner bailey covers 3/4 of an acre. The value of the
manor had risen since the Conquest, but it was only a small one, with
no villages in its soke.

RICHARD'S CASTLE, Herefordshire (Fig. 27).--There can be little doubt
that this is the castle referred to in Domesday Book under the name
of Avreton, as it is not far from Overton, on the northern border of
Hereford.[599] Richard's Castle is almost certainly the castle of
Richard, son of Scrob, one of the Normans to whom Edward the Confessor
had granted large estates, and who probably fortified himself on this
site. At the time of the Survey Richard was dead, and the castle was
held by his son Osbern, and it is noted that he pays 10s., but the
castle is worth 20s. to him. Its value was the same as in King Edward's
time, a fact worth noting, as it coincides with the assumption that
this was a pre-Conquest castle. There is a high and steep motte at
Richard's Castle, and a small half-moon shaped bailey.[600] There are
remains of a stone wing wall running down the motte, and on the top
there is a straight piece of masonry which must be part of a tower
keep. The area of the inner bailey is 2/3 of an acre. Avreton was
not the centre of a soke, but appears to have lain in the manor of
Ludeford.

[Illustration: FIG. 27. Rayleigh, Essex. Richard's Castle, Hereford.]

RICHMOND, Yorks (Fig. 28).--As in the case of Pontefract, this other
great Yorkshire castle is not mentioned by name in Domesday Book,
nor is there any allusion to it except a casual mention in the
_Recapitulation_ that Earl Alan has 199 manors in his castelry, and
that besides the castelry he has 43 manors.[601] The castle must have
been built at the date of the Survey, which was completed only a year
before William I.'s death; for during William's lifetime Earl Alan, the
first holder of the fief, gave _the chapel in the castle of Richmond_
to the abbey of St Mary at York, which he had founded.[602] The name,
of course, is French, and it seems impossible now to discover what
English manor-name it has displaced.[603] It is certainly a case in
which the Norman castle was not placed in the seat of the former Saxon
proprietor, but in the site which seemed most defensible to the Norman
lord. The lands of Earl Alan in the wapentake of Gilling had belonged
to the Saxon Earl Edwin, and thus cannot have fallen to Alan's share
before Edwin's death in 1071. The _Genealogia_ published by Dodsworth
(from an MS. compiled in the reign of Edward III.), says that Earl
Alan first built Richmond Castle near his chief manor of Gilling, to
defend his people against the attacks of the disinherited English and
Danes.[604] The passage has been enlarged by Camden, who says that
Alan "thought himself not safe enough in Gilling"; and this has been
interpreted to mean that Alan originally built his castle at Gilling,
and afterwards removed it to Richmond; but the original words have no
such meaning.[605]

Richmond Castle differs from most of the castles mentioned in Domesday
in that it has no motte. The ground plan indeed was very like that
of a motte-and-bailey castle, in that old maps show a small roundish
enclosure at the apex of the large triangular bailey.[606] But a recent
examination of the keep by Messrs Hope and Brakespear has confirmed
the theory first enunciated by Mr Loftus Brock,[607] that the keep
is built over the original gateway of the castle, and that the lower
stage of its front wall is the ancient wall of the castle. The small
ward indicated in the old maps is therefore most likely a barbican,
of later date than the 12th century keep, which is probably rightly
attributed by the _Genealogia_ cited above to Earl Conan, who reigned
from 1148-1171.[608] Some entries in the _Pipe Rolls_ make it almost
certain that it was finished by Henry II., who kept the castle in
his own hands for some time after the death of Conan.[609] There are
some indications at Richmond that the first castle was of stone and not
of earth and wood. The walls do not stand on earthen banks; the Norman
curtain can still be traced on two sides of the castle, and on the
west side it seems of early construction, containing a great deal of
herringbone work, and might possibly be the work of Earl Alan.

[Illustration: FIG. 28. Richmond, Yorks. Rochester, Kent.]

The whole area of the castle is 2-1/2 acres, including the annexe known
as the Cockpit. This was certainly enclosed during the Norman period,
as it has a Norman gateway in its wall.

As we do not know the name of the site of Richmond before the Conquest,
and as the name of Richmond is not mentioned in Domesday Book, we
cannot tell whether the value of the manor had risen or fallen. But no
part of Yorkshire was more flourishing at the time of the Survey than
this wapentake of Gilling, which belonged to Earl Alan; in no district,
except in the immediate neighbourhood of York, are there so many places
where the value has risen. Yet the greater part of it was let out to
under-tenants.

ROCHESTER, Kent (Fig. 28).--Under the heading of Aylsford, Kent, the
Survey tells us that "the bishop of Rochester holds as much of this
land as is worth 17s. 4d. _in exchange for the land in which the castle
sits_."[610] Rochester was a Roman _castrum_, and portions of its Roman
wall have recently been found.[611] The fact that various old charters
speak of the _castellum_ of Rochester has led some authorities to
believe that there was a castle there in Saxon times, but the context
of these charters shows plainly that the words _castellum Roffense_
were equivalent to _Castrum Roffense_ or _Hrofesceastre_.[612]
Otherwise there is not a particle of evidence for the existence of a
castle at Rochester in pre-Norman times, and the passage in Domesday
quoted above shows that William's castle was a new erection, built on
land obtained by exchange from the church.

Outside the line of the Roman wall, to the south of the city, and west
of the south gate, there is a district called Boley or Bullie Hill,
which at one time was included in the fortifications of the present
castle. It is a continuation of the ridge on which that castle stands,
and has been separated from it by a ditch. This ditch once entirely
surrounded it, and though it was partly filled up in the 18th century
its line can still be traced. The area enclosed by this ditch was
about 3 acres; the form appears to have been oblong. In the grounds
of Satis House, one of the villas which have been built on this site,
there still remains a conical artificial mound, much reduced in size,
as it has been converted into a pleasure-ground with winding walks,
but the retaining walls of these walks are composed of old materials;
and towards the riverside there are still vestiges of an ancient
wall.[613] We venture to think that this Boley Hill and its motte
formed the original site of the (probably) wooden castle of William
the Conqueror. Its nature, position, and size correspond to what we
have already observed as characteristic of the first castles of the
Conquest. It stands on land which originally belonged to the church of
St Andrew, as Domesday Book tells us William's castle did.[614] The
very name may be interpreted in favour of this theory.[615] And that
there was no Roman or Saxon fortification on the spot is proved by
excavations, which have shown that both a Roman and a Saxon cemetery
occupied portions of the area.[616]

It is well known that between the years 1087 and 1089 the celebrated
architect, Gundulf, Bishop of Rochester, built a new _stone_ castle
for William Rufus, "in the best part of the city of Rochester."[617]
This castle, of course, was on the same site as the present one,
though the splendid keep was not built till the next reign.[618] But
if what we have maintained above be correct the castle of Gundulf was
built on a different site from that of the castle of William. Nor are
we without evidence in support of this. What remains of the original
Norman wall of Gundulf's castle (and enough remains to show that the
circuit was complete in Norman times) does not stand on earthen banks;
and this, though not a proof, is a strong suggestion that there was
no earthen bank belonging to some previous castle when Gundulf began
his building.[619] But further, Mr Livett has shown in his paper on
_Mediæval Rochester_[620] that in order to form a level plateau for
the court of the castle the ground had to be artificially made up on
the north and east sides, and in these places the wall rests on a
foundation of gravel, which has been forcibly rammed to make it solid,
and which goes through the artificial soil to the natural chalk below.
Now what can this rammed gravel mean but an expedient to avoid the
danger of building in stone on freshly heaped soil? Had the artificial
platform been in existence ever since the Conquest, it would have been
solid enough to build upon without this expense. It is therefore at
least probable that Bishop Gundulf's castle was built on an entirely
new site.

It seems also to be clear that the Boley Hill was included as an
outwork in Bishop Gundulf's plan, for the castle ditch is cut through
the Roman wall near the south gate of the city.[621] Mr Livett remarks
that King John appears to have used the hill as a point of vantage
when he attacked the city in 1215, and he thinks this was probably the
reason why Henry III.'s engineers enclosed it with a stone wall when
they restored the walls of the city.[622] Henry III.'s wall has been
traced all round the city, and at the second south gate it turns at
right angles, or nearly so, so as to enclose Boley Hill.[623] It is
probable, as Mr Livett suggests, that the drawbridge and _bretasche_,
or wooden tower, ordered in 1226 for the southern side of Rochester
Castle,[624] were intended to connect the Boley Hill court with the
main castle. In 1722 the owner of the castle (which had then fallen
into private hands) conveyed to one Philip Brooke, "that part of the
castle ditch and ground, as it then lay unenclosed, on Bully Hill,
being the whole breadth of the hill and ditch without the walls of the
castle, extending from thence to the river Medway."[625]

The general opinion about the Boley Hill is that it is a Danish
earthwork, thrown up by the Danes when they besieged the city in 885.
But if our contention in Chapter IV. is just, the Danish fortifications
were not mottes, nor anything like them; and (as has already been
pointed out) the _Anglo-Saxon Chronicle_ indicates the nature of the
fortress in this case by its expression, "they made a work around
themselves";[626] that is, it was a circumvallation. Moreover, at
Rochester the Danes would have had to pass under the bridge (which is
known to have existed both in Roman and Saxon times) in order to get to
the Boley Hill; and even if their ships were small enough to do this
they would hardly have been so foolish as to leave a bridge in their
possible line of retreat. It is therefore far more likely that their
fastness was somewhere to the north or east of the city.[627]

It is a noteworthy fact that up till very recently the Boley Hill had
a special jurisdiction of its own, under an officer called the Baron
of the Bully, appointed by the Recorder of the city. This appears to
date from a charter of Edward IV. in 1460, which confirms the former
liberties of the citizens of Rochester, and ordains that they should
keep two courts' leet and a court of pie-powder annually on the Bullie
Hill. The anonymous historian of Rochester remarks that it was thought
that the baron represented the first officer under the governor of the
castle before the court leet was instituted, to whose care the security
of the Bullie Hill was entrusted.[628] This is probably much nearer the
truth than the theory which would assign such thoroughly feudal courts
as those of court leet and pie-powder to an imaginary community of
Danes residing on the Boley Hill. When we compare the case of the Boley
Hill with the somewhat similar cases of Chester and Norwich castles
we shall see that what took place in Edward IV.'s reign was probably
this: the separate jurisdiction which had once belonged to an abandoned
castle site was transferred to the citizens of Rochester, but with the
usual conservatism of mediæval legislation, it was not absorbed in the
jurisdiction of the city.

The value of Rochester at the time of the Survey had risen from 100_s._
to 20_l._[629] The increase of trade, arising from the security of
traffic which was provided by William's castles on this important
route, no doubt accounts in great measure for this remarkable rise in
value.

ROCKINGHAM, Northants (Fig. 29).--Here, also, the castle was clearly
new in William's reign, as the manor was uninhabited (_wasta_) until a
castle was built there by his orders, in consequence of which the manor
produced a small revenue at the time of the Survey.[630] The motte,
now in great part destroyed, was a large one, being about 80 feet
in diameter at the top; attached to it is a bailey of irregular but
rectilateral shape (determined by the ground) covering about 3 acres.
There is another large bailey to the S. covering 4 acres, formed by
cutting a ditch across the spur of the hill on which the castle stands,
which is probably later. The first castle would undoubtedly be of wood,
and it is probable that King John was the builder of the "exceeding
fair and strong" keep which stood on the motte in Leland's time,[631]
as there is an entry in the _Pipe Roll_ of the thirteenth year of his
reign for 126_l._ 18_s._ 6_d._ for the work of the new tower.[632] This
keep, if Mr Clark is correct, was polygonal, with a timber stockade
surrounding it.

Rockingham was only a small manor of one hide in Saxon times, though
its Saxon owner had sac and soke. It stands in a forest district, not
near any of the great ancient lines of road, and was probably built for
a hunting seat.

The value of the manor had risen at the time of the Survey.[633]

During the Civil War, the motte of Rockingham was fortified in an
elaborate manner by the Parliamentarians, part of the defences being
two wooden stockades:[634] an interesting instance of the use both of
mottes and of wooden fortifications in comparatively modern warfare.
Only the north and west sides of this mount now remain.

[Illustration: FIG. 29. Rockingham, Northants.]

OLD SARUM, Wilts (Fig. 30).--Sir Richard Colt Hoare printed in his
_Ancient Wiltshire_ a document purporting to be an order from Alfred,
"King of the English," to Leofric, "Earl of Wiltunshire," to maintain
the castle of Sarum, and add another ditch to it.[635] The phraseology
of the document suggests some doubts of its genuineness, and though
there would be nothing improbable in the theory that Alfred reared
the outer bank of the fortress, recent excavations have shown that
the place was occupied by the Romans, and therefore make it certain
that its origin was very much earlier than Alfred's time. Moreover,
the convergence of several Roman roads at this spot suggests the
probability of a Roman station,[636] while the form of the enclosure
renders an earlier origin likely. Domesday Book does not speak of
Salisbury as a _burgus_, and when the _burgus_ of Old Sarum is
mentioned in later documents it appears to refer to a district lying at
the foot of the Castle Hill, and formerly enclosed with a wall.[637]
Nor is it one of the boroughs of the _Burghal Hidage_. But that Sarum
was an important place in Saxon times is clear from the fact that there
was a mint there; and there is evidence of the existence of at least
four Saxon churches, as well as a hospital for lepers.[638]

For more exact knowledge as to the history of this ancient fortress
we must wait till the excavations now going on are finished, but in
the meanwhile it seems probable that the theory adopted by General
Pitt-Rivers is correct. He regarded Old Sarum as a British earthwork,
with an inner castle and outer barbicans added by the Normans. After
building this castle in the midst of it the Normans appear to have
considered the outer and larger fortification too valuable to be
given up to the public, but retained it under the government of the
castellan, and treated it as part of the castle.

There is no mention of the castle of Salisbury in Domesday Book, but
the bishop is named as the owner of the manor.[639] The episcopal see
of Sherborne was transferred to Sarum in 1076 by Bishop Hermann, in
accordance with the policy adopted by William I. that episcopal sees
should be removed from villages to towns:[640] a measure which in
itself is a testimony to the importance of Salisbury at that time.
The first mention of the castle is in the charter of Bishop Osmund,
1091.[641] The bishop was allowed to lay the foundations of his new
cathedral within the ancient fortress. As might be expected, friction
soon arose between the castellans and the ecclesiastics; the castellans
claimed the custody of the gates, and sometimes barred the canons,
whose houses seem to have been outside the fortress, from access to
the church. These quarrels were ended eventually by the removal of the
cathedral to the new town of Salisbury at the foot of the hill.

[Illustration: FIG. 30. Old Sarum, Wilts.]

The position of the motte of Old Sarum is exceptional, as it stands in
the centre of the outer fortress. This must be owing to the position
of the ancient vallum, encircling the summit of one of those round,
gradually sloping hills so common in the chalk ranges, which made it
necessary to place the motte in the centre, because it was the highest
part of the ground. The present excavations have shown that it is in
part artificial. But though the citadel was thus exceptionally placed,
the principle that communication with the outside must be maintained
was carried out; the motte had its own bailey, reaching to the outer
vallum. The remains of three cross banks still exist, two of which must
have enclosed the _magnum ballium_ which is spoken of in the _Pipe
Rolls_ of Henry II. Probably this bailey occupied the south-eastern
third of the circle, which included the main gateway and the road to
the citadel. In the ditch on the north side of this enclosure, an
arched passage, apparently of Norman construction, was found in 1795;
it was doubtless a postern or sallyport.[642] The main entrance is
defended by a separate mount with its own ditch, which is conjectured
to be of later date than the vallum itself. The area of the top of the
motte is about 1-3/4 acres, a larger size than usual, but not larger
than that of several other important castles.[643] In Leland's time
there was "much notable ruinous building" still remaining of this
fortress, and the excavations have already revealed the lower portions
of some splendid walls and gateways, and the basement of a late Norman
keep which presents some unusual features.[644] The earthworks,
however, bear witness to a former wooden stockade both to the citadel
and the outer enclosure. The top of the motte is still surrounded by
high earthen banks.

As that great building bishop, Roger of Salisbury (1099-1139), is
said to have environed the castle with a new wall,[645] it would
seem likely that he was the first to transform the castle from wood
to stone. But in Henry II.'s reign, we find an entry in the _Pipe
Rolls_ for materials for enclosing the great bailey. An order for the
destruction of the castle had been issued by Stephen,[646] but it is
doubtful whether it was carried out. The sums spent by Henry II. on the
castle do not amount to more than £266, 12s. 5d., but the work recently
excavated which appears to be of his date is very extensive indeed.

The mention of a small wooden tower in Richard I.'s reign shows that
some parts of the defences were still of wood at that date.[647]
Timber and rods for _hoarding_ the castle, that is, for the wooden
machicolations placed at the tops of towers and walls, were ordered at
the end of John's reign.[648]

It is not known when the castle was abandoned, but the list of
castellans ceases in the reign of Henry VI., when it was granted to
the Stourton family.[649] Though the earls of Salisbury were generally
the custodians of Sarum Castle, except in the time of Bishop Roger, it
was always considered a royal castle, while the manor belonged to the
bishop.[650] It is remarked in the _Hundred Rolls_ of Henry III., that
no one holds fiefs for ward in this castle, and that nothing belonged
to the castle outside the gate.[651]

The value of the manor of Salisbury appears to have risen very greatly
since the Conquest.[652]

SHREWSBURY (Fig. 31).--The passage in Domesday Book relating to this
town has been called by Mr Round one of the most important in the
Survey, and it is of special importance for our present purpose. "The
English burghers of Shrewsbury say that it is very grievous to them
that they have to pay all the geld which they paid in King Edward's
time, although the castle of the earl occupies [the site of] 51 houses,
and another 50 are uninhabited."[653] It is incomprehensible how in the
face of such a clear statement as this, that the new castle occupied
the site of fifty-one houses, anyone should be found gravely to
maintain that the motte at Shrewsbury was an English work; for if the
motte stood there before, what was the clearance of houses made for?
The only answer could be to enlarge the bailey. But this is exactly
what the Norman would not wish to do; he would want only a small
area for the small force at his disposal for defence. Shrewsbury was
certainly a borough (that is, a fortified town) in Anglo-Saxon times;
probably it was one of the towns fortified by Ethelfleda, though it
is not mentioned by name in the list of those towns furnished by the
_Anglo-Saxon Chronicle_.[654] Its ancient walls were certainly only
of earth and wood, for a writ of 1231 says that the old stockade and
the old bretasche of the old ditch of the town of Shrewsbury are to be
granted to the burghers for strengthening the new ditch.[655]

The castle of Shrewsbury was built on the neck of the peninsula on
which the town stands, and on the line of the town walls. The oval
motte, which still remains, stands, as usual, on the line of the castle
banks, and slopes steeply down to the Severn on one side. Its nearness
to the river made it liable to damage by floods. Thus we find Henry
II. spending 5_l._ on the repair of the motte,[656] and in Edward I.'s
reign the abbot's mill is accused of having caused damage to the extent
of 60 marks to the motte. But the men of the hundred exonerate the
mill, and from another passage the blame appears to lie on the fall of
a great wooden tower.[657] This can hardly have been other than the
wooden keep on the motte, and thus we learn the interesting fact that
as late as Edward I.'s reign the castle of Shrewsbury had only a
wooden keep. The present tower on the motte is the work of Telford.

[Illustration: FIG. 31. Shrewsbury. Skipsea, Yorks.]

The bailey of Shrewsbury Castle is roughly semilunar and covers nearly
an acre. The walls stand on banks, which shows that the first wall was
of timber. The Norman entrance arch seems to render it probable that it
was in Henry II.'s reign that stone walls were first substituted for a
wooden stockade, and the _Pipe Rolls_ contain several entries of sums
spent by Henry on this castle.[658] But the first mention of stone in
connection with the castle is in the reign of Henry III.[659] In the
reign of Edward I., a _jarola_ or wooden wall, which had been raised
above the outer ditch in the time of the Barons' War, was replaced
by a stone wall.[660] This perhaps refers to the second bailey, now
destroyed, which lay to the south of the castle. In the time of Charles
I. the castle still had a wooden palisade on the counterscarp of the
ditch.[661] The two large drum towers on the walls, and the building
between them, now converted into a modern house, belong to a much later
period than the walls. The area of the present castle, including the
motte, is 4/5 of an acre.

The value of the town of Shrewsbury had risen since the Conquest.

SKIPSEA, Yorks (Fig. 31).--There is no mention of this castle in
Domesday Book, but the chronicle of Meaux Abbey tells us that it
was built by Drogo de Bevrère in the reign of William I.[662] This
chronicle is not indeed contemporary, but its most recent editor
regards it as based on some much earlier document. It was the key of
the great manor of Holderness, which the Conqueror had given to Drogo,
but which Drogo forfeited by murdering his wife, probably on this very
site. The situation of Skipsea is remarkable, but the original plan of
Kenilworth Castle presented a close parallel to it. The motte, which
is 46 feet high, and 1/5 of an acre in space on top, is separated from
the bailey by a level space, which was formerly the Mere of Skipsea,
mentioned in documents of the 13th century, which reckon the take of
eels in this mere as a source of revenue.[663] The motte thus formed
an island in the mere, but as an additional defence--perhaps when the
mere began to get shallow--it was surrounded by a bank and ditch of its
own. No masonry is to be seen on the motte now, except a portion of a
wing wall going down it. It is connected with its bailey on the other
side of the mere by a causeway which still exists. This bailey is of
very unusual size, covering 8-1/4 acres; its banks still retain the
name of the Baile Welts, and one of the entrances is called the Baile
Gate. Skipsea Brough, which no doubt represents the former _burgus_
of Skipsea, is outside this enclosure, and has no defences of its
own remaining. A mandate of Henry III. in 1221, ordered the complete
destruction of this castle,[664] and it was no doubt after this that
the earls of Albemarle, who had succeeded to Drogo's estates, removed
their _caput baroniæ_ to Burstwick.[665]

The value of the manor of Cleeton, in which Skipsea lies, had fallen at
Domesday.[666]

STAFFORD (Fig. 32).--The _Anglo-Saxon Chronicle_ says that Ethelfleda
of Mercia built the _burh_ of Stafford; and consequently we find that
both in King Edward and King William's time Stafford was a burgus,
or fortified town. Florence of Worcester, who is considered to have
used a superior copy of the _Chronicle_ as the foundation of his work,
says that Ethelfleda built an _arx_ on the north bank of the Sowe
in 914. _Arx_, in our earlier chronicles, is often only a bombastic
expression for a walled town, as, for example, when Ethelwerd says
that Ethelfleda's body was buried in St Peter's porch in the _arx_
of Gloucester.[667] But the statement led many later writers, such
as Camden, to imagine that Ethelfleda built a _tower_ in the town of
Stafford; and these imaginings have created such a tangled skein of
mistake that we must bespeak our readers' patience while we attempt to
unravel it.

Domesday Book only mentions Stafford Castle under the manor of Chebsey,
a possession of Henry de Ferrers. Its words are: "To this manor
_belonged_ the land of Stafford, in which the king commanded a castle
to be built, which is now destroyed."[668] Ordericus also says that the
king placed a castle at Stafford, on his return from his third visit to
the north, in 1070.[669] Now the language of Domesday appears to us to
say very plainly that in the manorial rearrangement which followed the
Conquest some land was taken out of the manor of Chebsey, which lies
immediately to the south of the borough of Stafford, to furnish a site
for a royal castle.[670] It is exactly in this position that we now
find a large oblong motte, similar to the other mottes of the Conquest,
and having the usual bailey attached to it. It lies about a mile and a
half south-west of the town, near the main road leading into Shropshire.

The position was an important one, as the castles of Staffordshire
formed a second line of defence against the North Welsh, as well as
a check to the great palatinate earls of Shropshire.[671] The motte
itself stood on high ground, commanding a view of twenty or thirty
miles round, and both Tutbury and Caus castles could be seen from
it. Between it and the town lies a stretch of flat ground which has
evidently been a swamp formerly, and which explains the distance of the
castle from the town; while the fact that it lies to the _south_ of the
Sowe shows that it has no connection with Ethelfleda's work. There is
no dispute that this motte was the site of the later baronial castle of
Stafford, the castle besieged and taken in the Civil War; the point we
have to prove is that it was also the castle of Domesday Book.[672]

[Illustration: FIG. 32. Stafford. Stanton Holgate, Salop. Tamworth,
Staffs. Tickhill, Yorks.]

If the first castle of Stafford was of earth and wood, like most of
William's castles, there would be nothing wonderful in its having
many destructions and many resurrections. This castle was clearly a
royal castle, from the language of Domesday Book. As a royal castle
it would be committed to the custody of the sheriff, who appears to
have been Robert de Stafford,[673] ancestor of the later barons of
Stafford, and brother of Ralph de Todeni, one of the great nobles
of the Conquest. Ralph joined the party of Robert Curthose against
Henry I. in 1101, and it is conjectured that his brother Robert was
involved in the same rebellion, for in that year we find the castle
held for the king by William Pantolf, a trusty companion of the
Conqueror.[674] It is very unlikely that this second castle of Stafford
was on a different site from the one which had been destroyed; and an
ingenious conjecture of Mr Mazzinghi's helps us to identify it with
the castle on the motte. In that castle, when it again emerges into
light in the reign of Henry II., we find a chapel dedicated to St
Nicholas, which Robert de Stafford gives to the abbey of Stone, and the
king confirms the gift.[675] The worship of St Nicholas came greatly
into fashion after the translation of his remains from Asia Minor to
Bari, in Italy, in 1087. William Pantolf visited the shrine at Bari,
got possession of some of the relics of St Nicholas, and with great
reverence deposited them in his own church of Noron, in Normandy.[676]
It is therefore extremely probable that Pantolf founded the chapel
of St Nicholas in Stafford Castle during the time that the castle
was in his custody.[677] But about the situation of the chapel of St
Nicholas there is no doubt, as its history is traceable down to the
16th century. It stood in the bailey of the castle outside the town.
This castle was therefore certainly identical with that of Henry II.,
and most probably with that of Henry I. and William I.

So far, as we have seen, Stafford Castle was a royal castle. It is true
that in the reign of Henry II.'s predecessor, Stephen, we find the
castle again in the hands of a Robert de Stafford, who speaks of it
as "castellum meum."[678] Apparently the troubles of Stephen's reign
afforded an opportunity to the family of the first Norman sheriff to
get the castle again into their hands. But under the stronger rule
of Henry II. the crown recovered its rights, and the gift of the
chapel in the castle evidently could not be made without the consent
of the king. The gaol which Henry II. caused to be made in Stafford
was doubtless in this castle.[679] John repaired the castle,[680]
and ordered _bretasches_, or wooden towers, to be made in the forest
of Arundel, and sent to Stafford:[681] a statement which gives us an
insight into the nature of the castle in John's reign. But it was
the tendency of sheriffdoms to become hereditary, as Dr Stubbs has
pointed out,[682] and this seems to have been the case at Stafford. In
the reign of Edward I. a local jury decided that Nicholas, Baron of
Stafford, held the castle of Stafford from the king _in capite_, by the
service of three and a half knights' fees;[683] and in 1348, Ralph,
Baron of Stafford, obtained a license from Edward III. "to fortify and
crenellate his _manses_ of Stafford and Madlee with a wall of stone and
lime, and to make castles thereof."[684] The indenture made with the
mason a year previously is still extant, and states that the castle
is to be built upon the _moële_ in the manor, whereby the motte is
evidently meant.[685] Besides, the deed is dated "at the Chastel of
Stafford," showing that the new castle of stone and lime was on the
site of an already existing castle.

We might spin out further evidence of the identity of the site of
William's castle with that of the present one, from the name of the
manor of Castel, which grew up around it, displacing the equally
suggestive name of Montville, which we find in Domesday Book.[686]
Against the existence of another castle in the town we have the absence
of any such castle in William Smith's plan of 1588; the silence of
Speed and Leland, who only mention the present castle;[687] and the
statement of Plot, who wrote about the end of the 17th century,
that "he could not hear any footsteps remaining" of a castle in
Stafford.[688] We may therefore safely conclude that it was only due to
the fancy of some Elizabethan antiquary that in an old map of that time
a spot to the south-west of the town is marked with the inscription,
"The old castle, built by Edward the Elder, and in memorie fortified
with reel walls."[689]

The value of Stafford town had risen at the time of the Survey, as the
king had 7_l._ for his share, which would make the whole revenue to
king and earl 10_l._ 10_s._, as against 9_l._ before the Conquest. The
property of the canons of Stafford had risen from £1 to £3.[690]

The area of the bailey is 1-3/5 acres.

STAMFORD, Lincoln and Northants.--This was one of the boroughs
fortified by Edward the Elder, and consequently we find it a royal
_burgus_ at the time of the Survey. But Edward's borough, the
_Chronicle_ tells us, was on the south side of the Welland; the
northern borough, on the other side, may have been the work of the
Danes, as Stamford was one of the towns of the Danish confederacy of
the Five Boroughs. The Norman castle and its motte are on the north
side, and five _mansiones_ were destroyed for the site.[691] There is
at present no appearance of masonry on the motte, which is partly cut
away, and what remains of the castle wall is of the 13th century. It
is therefore probable that the _turris_, or keep, which surrendered to
Henry II. in 1153, was of wood.[692] Henry gave the castle to Richard
Humet, constable of Normandy, in 1155.[693] It was a very exceptional
thing that Henry should thus alienate a royal castle, and special
circumstances must have moved him to this act. The castle was destroyed
in Richard III.'s time, and the materials given to the convent of the
Carmelite Friars. It appears to have been within the town walls, with
a bailey stretching down to the river; this bailey is quadrangular.
An inquisition of 1341 states that "the site of the castle contains 2
acres."[694]

Stamford had risen enormously in value since the Conquest. "In King
Edward's time it paid 15_l._; now, it pays for _feorm_ 50_l._, and for
the whole of the king's dues it now pays 28_l._"[695]

STANTON, Stanton Long, in Shropshire (Fig. 32).--At the time of the
Survey, the Norman Helgot was Lord of Corve Dale, and had his castle
at Stanton.[696] The castle was afterwards known as Helgot's Castle,
corrupted into Castle Holdgate. The site has been much altered by the
building of a farmhouse in the bailey, but the motte still exists,
high and steep, with a ditch round about half its circumference;
there are some traces of masonry on the top. One side of the bailey
ditch is still visible, and a mural tower of Edwardian style has
been incorporated with the farmhouse. The exact area cannot now be
calculated, but it can hardly have exceeded 2-1/2 acres. The manor of
Stanton was an agglomeration of four small manors which had been held
by different proprietors in Saxon times, so it was not the centre of a
soke. The value of the manor had risen.

TAMWORTH, Stafford (Fig. 32).--Although Tamworth Castle is not
mentioned in Domesday Book, it must have been in existence in the
11th century, as a charter of the Empress Matilda mentions that
Robert le Despenser, brother of Urso d'Abetot, had formerly held this
castle;[697] now Urso d'Abetot was a contemporary of the Conqueror,
and so must his brother have been. Tamworth Castle stands on a motte
50 feet high, and 100 feet in diameter across the top, according to
Mr Clark. It is an interesting instance of what is commonly called a
shell keep, with a stone tower; one of the instances which suggest that
the shell did not belong to a different type of castle to the tower,
but was simply a ward wall, which probably at first enclosed a wooden
tower. The tower and wall (or chemise) are probably late Norman, but
the remarkable wing wall (there is only one, instead of the usual two)
which runs down the motte is entirely of herring-bone work, and _may_
be as old as Henry I.'s time.[698] A bailey court, which cannot have
been large, lay between the motte and the river Tame, but its outline
cannot now be determined, owing to the encroachments of buildings.
Tamworth is about a mile from the great Roman road known as Watling
Street. We have already referred to the fortification of the _burh_
here by Ethelfleda;[699] probably she only restored walls or banks
which had existed before round this ancient capital of Mercia.

The value of the manor of Tamworth is not given in Domesday Book.

TICKHILL, Yorks (Fig. 32).--The name Tickhill does not occur in
Domesday, but it is covered by that of Dadesley, the manor in which
this castle was built: a name which appears to have gone out of use
when the _hill_ was thrown up. There can be no doubt that it was the
castle of Roger de Busli, one of the most richly endowed of William's
tenants-in-chief, as it is mentioned as such by Ordericus.[700] He
calls it the castle of Blythe, a name which it probably received
because Blythe was the most important place near, and Dadesley was so
insignificant. Florence of Worcester, when describing the same events,
calls the castle Tykehill. The remains furnish an excellent specimen of
the earthworks of this class. The motte is 75 feet high, and its area
on top about 80 feet in diameter; about a third of it is natural, the
rest artificial. Only a slight trace remains of the ditch separating
it from the oval bailey, which covers 2 acres. The foundations of a
decagonal tower, built in the reign of Henry II., are still to be seen
on the top.[701] The bailey retains its banks on the scarp, surmounted
now by a stone curtain, which, along with the older part of the
gatehouse, is possibly of the time of Henry I.[702] The outer ditch
is about 30 feet broad, and is still full of water in parts. On the
counterscarp a portion of the bank remains. This bank carried a wooden
palisade when the castle was besieged by Cromwell.[703] The site is
not naturally defensible; it is about three and a half miles from the
northern Roman road.

The value of the manor of Dadesley had risen at the time of the
Survey.[704] The stone buildings which once stood in the bailey have
been transformed into a modern house.

TONBRIDGE, Kent (Fig. 33).--This notable castle, the first English
seat of the powerful family who afterwards took their name from Clare
in Suffolk, is first mentioned in 1088, when it was stormed by William
Rufus and his English subjects, who had adopted his cause against the
supporters of his brother Robert.[705] The castle was one of great
importance at several crises in English history; but it began as a
wooden keep on a motte, and the stone shell which now crowns this motte
cannot be earlier than the 12th century, and judging by its buttresses,
is much later. The castle stands outside the town of Tonbridge,
separated from it by moats which were fed from the river. The smaller
bailey of 1-1/2 acres, probably the original one, is square, with
rounded corners. The palatial gatehouse, of the 13th or 14th century,
is a marked feature of this castle. There appears to have been only
one wing wall down the motte to the bailey, but a second one was not
needed, owing to the position of the motte with regard to the river.

The value of the manor of Hadlow, in which Tonbridge lay, was
stationary at Domesday.[706] It belonged to the see of Canterbury, and
was held by Richard de Bienfaite, ancestor of the House of Clare, as
a tenant of the see.

[Illustration: FIG. 33. Tonbridge, Kent. Totnes, Devon.]

TOTNES, Devonshire (Fig. 33).--The castle of Totnes belonged to
Judhael, one of King William's men, who has been already mentioned
under Barnstaple. This castle is not noticed in Domesday Book, but
its existence in the 11th century is made certain by a charter of
Judhael's giving land _below his castle_ to the Benedictine priory
which he had founded at Totnes: a charter certainly of the Conqueror's
reign, as it contains a prayer for the health of King William.[707]
The site was an important one; Totnes had been one of the boroughs of
the _Burghal Hidage_; it was at the head of a navigable river, and was
the point where the ancient Roman (?) road from Devonshire to Bath and
the North began its course.[708] The motte of the castle is very high
and precipitous, and has a shell on top, which is perfect up to the
battlements, and appears to be rather late Norman. This keep is entered
in a very unusual way, by a flight of steps leading up from the bailey,
deeply sunk in the upper part into the face of the motte, so as to
form a highly defensible passage. Two wing walls run down to the walls
of the bailey. There is at present no ditch between the motte and the
bailey. The whole area of the work is 3/4 acre. It stands in a very
defensible situation on a spur of hill overlooking the town, and lies
just outside the ancient walls.

The value of the town of Totnes had risen at Domesday.[709]

THE TOWER OF LONDON.--Here, as at Colchester, there is no motte,
because the original design was that there should be a stone
keep. Ordericus tells us that after the submission of London to
William the Conqueror he stayed for a few days in Barking while
certain fortifications in the city were being finished, to curb the
excitability of the huge and fierce population.[710] What these
fortifications were we shall never know, but we may imagine they were
earthworks of the usual Norman kind.[711] Certainly the great keep
familiarly known as the White Tower was not built in a few days;
it does not appear to have been even begun till some eleven years
later, when Gundulf, a monk celebrated for his architectural skill,
was appointed to the see of Rochester. Gundulf was the architect of
the Tower,[712] and it must therefore have been built during his
episcopate, which lasted from 1077-1108.[713] In 1097 we read that
"many shires which owe works to London were greatly oppressed in making
the wall (weall) round the Tower."[714] This does not necessarily mean
a stone wall, but probably it does, as Gundulf's tower can hardly have
been without a bank and palisade to its bailey.

As the Tower in its general plan represents the type of keep which was
the model for all succeeding stone keeps up to the end of the 12th
century, it seems appropriate here to give some description of its
main features. Its resemblance to the keep of Colchester, which also
was a work of William I.'s reign, is very striking.[715] Colchester is
the larger of the two, but the Tower exceeds in size all other English
keeps, measuring 118 × 98 feet at its base.[716] As it has been altered
or added to in every century, its details are peculiarly difficult to
trace, especially as the ordinary visitor is not allowed to make a
thorough examination.[717] Thus much, however, is certain: neither of
the two present entrances on the ground floor is original; the first
entrance was on the first floor, some 25 feet above the ground, at the
S.W. angle of the south side, and has been transformed into a window.
There was no entrance to the basement, but it was only reached by
the grand staircase, which is enclosed in a round turret at the N.E.
angle. There were two other stairs at the N.W. and S.W. angles, but
these only began on the first floor. The basement is divided by a cross
wall, which is carried up to the third storey. There are at present
three storeys above the basement. The basement, which is now vaulted
in brick, was not originally vaulted at all, except the south-eastern
chamber, under the crypt of the chapel.

The first floor, like the basement, is divided into three rooms, as,
in addition to the usual cross wall, the Tower has a branch cross wall
to its eastern section, which is carried up to the top. This floor
was formerly only lit by loopholes; Clark states that there were two
fireplaces in the east wall, but there is some doubt about this. The
S.E. room contained the crypt of the chapel, which was vaulted. It
is commonly supposed that the rooms on the first floor were occupied
by the guards of the keep. In the account which we have quoted from
Lambert of Ardres, the first floor is said to be the lord's habitation,
and the upper storey that of the guards; so that there seems to have
been no invariable rule.[718] No special room was allotted to the
kitchen, as in time of peace at any rate, the lord of the castle and
all his retainers took their meals in a great hall in the bailey of the
castle.[719] The ceilings of the two larger rooms of this floor are now
supported by posts, an arrangement which is probably modern, as the
present posts certainly are.[720]

The second floor contains the chapel, which in many keeps is merely
an oratory, but is here of unusual size. Its eastern end is carried
out in a round apse, a feature which is also found at Colchester, but
is not usual in Norman keeps.[721] It is a singularly fine specimen
of an early Norman chapel. This floor probably contained the royal
apartments; it was lighted by windows, not loops. Both the eastern and
western rooms had fireplaces; the eastern room goes by the name of the
Banqueting Chamber.

The third storey is on a level with the triforium of the chapel.[722]
This triforium is continued all round the keep as a mural passage, and
it has windows only slightly smaller than those of the floor below.
These mural galleries are found in most important keeps. As their
windows were of larger size than the loops which lit the lower floors,
it is possible that they may have been used for defence, either for
throwing down missiles or for shooting with bows and arrows. But no
near aim could be taken without a downward splay to the window, and
the bows of the 11th and 12th centuries were incapable of a long
aim. A plausible theory is that they were intended for the march of
sentinels.[723]

The masonry of the Tower is of Kentish rag, with ashlar quoins. In
mediæval times it had a forebuilding, with a round stair turret, which
is shown in some old views; but it may reasonably be doubted whether
this was an original feature.

As regards the ground plan of the castle as a whole, it is now
concentric, but was not so originally. The Tower was certainly placed
in the S.E. angle of the Roman walls of London, and very near the east
wall, portions of which have been discovered.[724] The conversion
of the castle into one of the concentric type was the work of later
centuries, and the history of its development has still to be
traced.[725]

TREMATON, Cornwall (Fig. 34).--"The Count [of Mortain] has a castle
there and a market, rendering 101 shillings."[726] Two Cornish castles
are mentioned in Domesday, and both of them are only on the borders
of that wild Keltic country; but while Launceston is inland, Trematon
guards an inlet on the south coast. The position of this castle is
extremely strong by nature, at the end of a high headland; on the
extreme point of this promontory the motte is placed. It carries a
well-preserved shell wall, which may be of Norman date, from the
plain round arch of the entrance.[727] It has been separated by a
ditch from the bailey, but the steepness of the hill rendered it
unnecessary to carry this ditch all round. The bailey, 1 acre in
extent, in which a modern house is situated, still has an entrance
gate of the 13th century, and part of a mediæval wall. A second
bailey, now a rose-garden, has been added at a later period. In spite
of the establishment of a castle and a market the value of the
manor of Trematon had gone down at the time of the Survey, which may
be accounted for by the fact that there were only ten ploughs where
there ought to have been twenty-four. It was only a small manor, and no
burgus is mentioned.

[Illustration: FIG. 34. Trematon, Cornwall. Tutbury, Staffs.]

TUTBURY, Staffordshire (Fig. 34).--In the magnificent earthworks of
this castle, and the strength of its site, we probably see a testimony
to the ability of Hugh d'Avranches; for we learn from Ordericus that in
1070 William I. gave to Henry de Ferrers the castle of Tutbury, which
had belonged to Hugh d'Avranches,[728] to whom the king then gave the
more dangerous but more honourable post of the earldom of Chester.
Domesday Book simply states that Henry de Ferrers has the castle of
Tutbury, and that there are forty-two men living by their merchandise
alone in the borough round the castle.[729]

At Tutbury the keep was placed on an artificial motte, which itself
stood on a hill of natural rock, defended on the N.W. side by
precipices. There is no trace of any ditch between the motte and
bailey. At present there is only the ruin of a comparatively modern
tower on the motte, but Shaw states that there was formerly a stone
keep.[730] A description of Elizabeth's reign says, "The castle is
situated upon a round hill, and is circumvironed with a strong wall
of astilar [ashlar] stone.... The king's lodging therein is fair and
strong, bounded and knit to the wall. And a fair stage hall of timber,
of a great length. Four chambers of timber, and other houses well
upholden, within the walls of the castle."[731] The king's lodging
will no doubt be the closed gatehouse; the custom of erecting gatehouse
palaces arose as early as the 13th century. This account shows how many
of the castle buildings were still of timber in Elizabeth's reign.

The bailey is quadrant-shaped, and has the motte at its apex. Its area
is 2-1/2 acres. Its most remarkable feature is that it still retains
its ancient banks on the east side and part of the south, and the
more recent curtain is carried on top of them. This curtain is of the
same masonry as the three remaining towers, which are of excellent
Perpendicular work, and are generally attributed to John of Gaunt,
who held this castle after his marriage with Blanche of Lancaster.
The first castle was undoubtedly of wood; it was pulled down by
order of Henry I. in 1175,[732] nor does there seem to have been any
resurrection till the time of Earl Thomas of Lancaster at the earliest.

Though Tutbury was the centre of the Honour of Ferrers, it does not
seem to have been even a manor in Saxon times. The borough was probably
the creation of the castellan, who also founded the Priory.[733] There
is no statement in the Survey from which we can learn the value T. R.
E., but T. R. W. it was 4_l._ 10_s._

TYNEMOUTH, Northumberland.--Besieged and taken by William Rufus in
1095.[734] There is no motte there, and probably never was one, as the
situation is defended by precipitous cliffs on all sides but one, where
a deep ditch has been cut across the neck of the headland.

[Illustration: FIG. 35. Wallingford, Berks.]

WALLINGFORD, Berkshire (Fig. 35).--There is good reason to suppose
that in the _vallum_ of the town of Wallingford we have an interesting
relic of Saxon times. Wallingford is one of the boroughs enumerated
in the _Burghal Hidage_; it was undoubtedly a fortified town at the
time of the Conquest,[735] and is called a _burgus_ in Domesday
Book; but there appears to be no evidence to connect it with Roman
times except the discovery of a number of Roman coins in the town
and its neighbourhood. No Roman buildings or pavements have ever
been found.[736] The Saxon borough was built on the model of a Roman
_chester_: a square with rounded corners. The rampart of Wallingford,
which still exists in great part, is entirely of earth, and must
have been crowned with a wooden wall, such as was still existing at
Portsmouth in Leland's time.[737] The accounts of Wallingford in the
great Survey are very full and important. "King Edward had eight
virgates in the borough of Wallingford, and in these there were 276
haughs paying 11_l._ of rent. Eight have been destroyed for the
castle."[738] This Norman castle was placed in the N.E. corner of
the borough. At present its precincts cover 30 acres,[739] but this
includes garden grounds, and no doubt represents later enclosures.
No ancient plan of the castle has been preserved, but from Leland's
description there appear to have been three wards in his time, each
defended by banks and ditches. The inner ward, which was doubtless
the original one, is rudely oblong in shape; it covers 4-1/2 acres.
Leland says, "All the goodly buildings, with the towers and dungeon,
be within the third dyke." The motte, which still exists, was on the
south-eastern edge of this ward; that is, it was so placed as to
overlook both the borough and the ford over the Thames.[740] It was
ditched around, and is said to have had a stone keep on the top; but
no foundations were found when it was recently excavated. It was found
to rest on a foundation of solid masonry several feet thick, sloping
upwards towards the outside, so that it must have stood in a kind of
stone saucer.[741] The masonry which remains in the other parts of the
castle is evidently none of it of the early Norman period, unless we
accept a fragment of wall which contains courses of tiles. Numerous
buildings were added in Henry III.'s reign; the walls and battlements
were repaired, and the _hurdicium_, which had been blown down by a high
wind, was renewed.[742] But the motte and the high banks show clearly
that the first Norman castle was of wood.

The value of the royal borough of Wallingford had considerably risen
since the Conquest.[743]

WARWICK (Fig. 36).--Here again we have a castle built on land which
the Conqueror obtained from a Saxon convent, a positive proof that
there was no castle there previously. Only a small number of houses
was destroyed for the castle,[744] and this points to the probability,
which is supported by some other evidence, that the castle was built
outside the town. Warwick, of course, was one of the boroughs fortified
by Ethelfleda, and it was doubtless erected to protect the Roman road
from Bath to Lincoln, the Foss Way, against the Danes. Domesday Book,
after mentioning that the king's barons have 112 houses in the borough,
and the abbot of Coventry 36, goes on to say that these houses belong
to the lands which the barons hold outside the city, and are rated
there.[745] This is one of the passages from which Professor Maitland
has concluded that the boroughs planted by Ethelfleda and her brother
were organised on a system of military defence, whereby the magnates
in the country were bound to keep houses in the towns.[746] Ordericus,
after the well-known passage in which he states that the lack of
castles in England was one great cause of its easy conquest by the
Normans, says: "The king _therefore_ founded a castle at Warwick, and
gave it in custody to Henry, son of Roger de Beaumont."[747] Putting
these various facts together, we may fairly assert that the motte
which still forms part of the castle of Warwick was the work of the
Conqueror, and not, as Mr Freeman believed, "a monument of the wisdom
and energy of the mighty daughter of Alfred,"[748] whose energy was
very much better employed in the protection of her people. Dugdale,
who also put the motte down to Ethelfleda, was only copying Rous, a
very imaginative writer of the 15th century.

The motte of Warwick is mentioned several times in the _Pipe Rolls_
of Henry II.; it then carried wooden structures on its top.[749] In
Leland's time there were still standing on this motte the ruins of a
keep, which he calls by its Norman name of the Dungeon. A fragment of
a polygonal shell wall still remains.[750] But there is not a scrap of
masonry of Norman date about the castle. The motte, and the earthen
bank which still runs along one side of the court, show that the first
castle was a wooden one. The bailey is oblong in shape, the motte being
outside it; its area is about 2-1/2 acres.

The value of Warwick had doubled since the Conquest.

[Illustration: FIG. 36. Warwick. Wigmore, Hereford.]

WIGMORE, Herefordshire (Fig. 36).--We have already referred to the
absurdity of identifying this place with the _Wigingamere_ of the
_Anglo-Saxon Chronicle_.[751] We have the strongest indication that the
Norman castle at Wigmore was a new erection, since Domesday Book tells
us that William FitzOsbern built it on waste land called Mereston.[752]
This express statement disposes of the fable in the _Fundationis
Historia_ of Wigmore Priory, that the castle of Wigmore had belonged
to Edric the Wild, and was rebuilt by Ralph Mortimer.[753] Wigmore had
only been a small manor of two taxable hides in Saxon times. Whereas
it had then been unproductive, at the date of the Survey there were two
ploughs in the demesne, and the borough attached to the castle yielded
7_l._ Here we have another instance of the planting of a borough close
to a castle, and of the revenue which was thus obtained.

There is a very large and high motte at Wigmore Castle, of oval shape,
on a headland which has been cut off by a deep ditch. The earthen
banks of its first fortification still remain, enclosing a small ward,
but on top of them is a wall in masonry, and the ruins of a polygonal
keep;[754] also the remains of two mural towers. Half-way down the end
of the headland, below the motte, is a small square court, which _may_
have been the original bailey; below it, again, is a larger half-moon
bailey furnished with walls and towers. But the whole area covered
is only 1 acre. The masonry is none of it earlier than the Decorated
period, except one tower in the bailey wall which may be late Norman.

[Illustration: FIG. 37. Winchester. (From a plan by W. Godson, 1750.)]

WINCHESTER, Hants.--We include Winchester among the castles mentioned
or alluded to in Domesday Book, because we think it can be proved that
the _domus regis_ mentioned under Alton and Clere is the castle built
by William outside the west gate of the city, where the present County
Hall is now almost the only remaining relic of any castle at all.[755]
Under the head of "Aulton" we are told that the abbot of Hyde had
unjustly gotten the manor in exchange for the king's house, because
by the testimony of the jurors it was already the king's house.[756]
That _excambio domus regis_ should read _excambio terræ domus regis_
is clear from the corresponding entry under Clere, where the words
are _pro excambio terræ in qua domus regis est in civitate_.[757] The
matter is put beyond a doubt by the confirmatory charter of Henry I. to
Hyde Abbey, where the king states that his father gave Aulton and Clere
to Hyde Abbey _in exchange for the land on which he built his hall in
the city of Winchester_.[758] Where, then, was this hall, which was
clearly new, since fresh land was obtained for it, and which must not
therefore be sought on the site of the palace of the Saxon kings? The
_Liber Winton_, a roll of Henry I.'s time, says that twelve burgesses'
houses had been destroyed and the land was now occupied by the king's
house.[759] Another passage says that a whole street _outside the west
gate_ was destroyed when the king made his ditch.[760] These passages
justify the conclusion of Mr Smirke that the king's house at Winchester
was neither more nor less than the castle which existed until the 17th
century outside the west gate.[761] Probably the reason why it is
spoken of so frequently in the earliest documents as the king's house
or hall, instead of the castle, is that in this important city, the
ancient capital of Wessex, where the king "wore his crown" once a
year, William built, besides the usual wooden keep on the motte, a
stone hall in the bailey, of size and dignity corresponding to the
new royalty.[762] In fact, the hall so magnificently transformed by
Henry III., and known to be the old hall of the castle, can be seen
on careful examination to have still its original Norman walls and
other traces of early Norman work.[763] The palace of the Saxon kings
stood, where we might expect to find the palace of native princes,
in the middle of the city; according to Milner it was on the site of
the present Square.[764] William may have repaired this palace, but
that he constructed two royal houses, a palace and a castle, is highly
improbable. The castle became the residence of the Norman kings, and
the Saxon palace appears to have been neglected.[765] We see with what
caution the Conqueror placed his castle at the royal city of Wessex
without the walls. Milner tells us that there was no access to it from
the city without passing through the west gate.[766] The motte of the
castle appears to have been standing in his time, as he speaks of "the
artificial mount on which the keep stands."[767] It is frequently
mentioned in mediæval documents as the _beumont_ or _beau mont_. It
was surrounded by its own ditch.[768] The bailey, if Speed's map is
correct, was triangular in shape. With its ditches and banks the castle
covered 6 acres, according to the commissioners who reported on it in
Elizabeth's reign; but the inner area cannot have been more than 4-1/2
acres. We may infer from the sums spent on this castle by Henry II.,
that he was the first to give it walls and towers of stone; the _Pipe
Rolls_ show entries to the amount of 1150_l._ during the course of his
reign; the work of the walls is frequently specified, and stone is
mentioned.

Domesday Book does not inform us whether the value of Winchester had
risen or fallen since the Conquest.

[Illustration: FIG. 38. Windsor Castle (From Ashmole's "Order of the
Garter.")]

WINDSOR (Fig. 38).--Here we have another of the interesting cases in
which the geld due from the tenant of a manor is lessened on account
of a castle having occupied a portion of the land.[769] The Survey
tells us that the castle of Windsor sits in half a hide belonging
to the manor of Clewer, which had become William's property as part
of the spoils of Harold. It was _now_ held of the king by a Norman
tenant-in-chief, but whereas it was formerly rated as five hides it
was now (that is, probably, since the castle was built) rated as four
and a half hides. Of course we are not to suppose that the castle
occupied the whole half hide, which might be some 60 acres; but it
extinguished the liability of that portion. At Windsor, however, we
have no occasion to press this argument as a proof that the castle was
new, since it is well established that the palace of the Saxon kings
was at least 2 miles from the present castle and town, in the village
long known as Old Windsor, which fell into decay as the town of Windsor
sprang up under the Norman castle.[770] The manor of Windsor was given
by Edward the Confessor to the convent of Westminster, but recovered by
the Conqueror.[771] But as the Survey shows us, he did not build his
castle in the manor of Windsor, but in that of Clewer. He built it for
a hunting-seat,[772] and it may have been for the purpose of recovering
forest rights that he resumed possession of Old Windsor; but he placed
his castle in the situation which he thought best for defence. For even
a hunting-seat in Norman times was virtually a castle, as many other
instances show.

It is needless to state that there is no masonry at Windsor of the
time of the Conqueror, or even of the time of his son Henry I., in
spite of the statement of Stowe that Henry "new builded the castle
of Windsor." This statement may perhaps be founded on a passage in
the _Anglo-Saxon Chronicle_ which says that Henry held his court for
the first time in the New Windsor in 1110. Perhaps the _Chronicle_
here refers to the _borough_ of New Windsor, as an entry in the _Pipe
Roll_ of Henry I. seems to show that he was the first to enclose the
_burgus_ of Windsor.[773] For it is probable that the first stone
castle at Windsor was built by Henry II., who spent £1670 on it in the
course of his reign. One of his first acts after his accession was an
exchange of land at Windsor, which seems to have been for the purpose
of a vineyard, and was possibly the origin of the second bailey.[774]
At present the position of the motte is central to the rest of the
castle, but this is so unusual that it suggests the idea that the upper
ward is the oldest, and that the motte stood on its outer edge. Henry
II. surrounded the castle with a wall, at a cost of about 128_l._[775]
The other entries in the _Pipe Rolls_ probably refer to the first
stone shell on the motte, and there is little doubt that the present
Round Tower, though its height has been raised in modern times, and
its masonry re-dressed and re-pointed so as to destroy all appearance
of antiquity, is in the main of Henry II.'s building. The frequent
payments for stone show the nature of Henry's work.

Although so much masonry was put up in Henry II.'s reign, the greater
part of what is now visible is not older than the time of Henry III.
The lower bailey seems to have been enlarged in his reign, as the
castle ditch was extended towards the town, and compensation given
for houses taken down.[776] The upper (probably the original) ward is
rectangular in shape, and with the motte and its ditches covers about
6-1/2 acres.[777] The state apartments, a chapel, and the Hall of
St George, are in the upper ward, showing that this was the site of
the original hall and chapel of the castle. The charter of agreement
between Stephen and Henry in 1153 speaks of the _motte_ of Windsor as
equivalent to the castle.[778] Repairs of the motte are mentioned in
the _Pipe Rolls_ of Henry II.[779]

The value of the manor of Clewer had fallen since the Conquest; that of
Windsor, which was worth 15_l._ T. R. E., but after the Conquest fell
to 7_l._, was again worth 15_l._ at the date of the Survey.[780]

WISBEACH, Cambridgeshire.--William I. built a castle here in 1072,
after suppressing the revolt of Hereward, in order to hold in check the
Cambridgeshire fen country.[781] There is an early mention of it in
the Register of Thorney Abbey. This castle, after being several times
rebuilt, is now completely destroyed, and "several rows of elegant
houses built on the site." Nevertheless, there still remain distinct
traces of the motte-and-bailey pattern in the gardens which now occupy
the site of the original castle of King William; the present Crescent
probably follows the line of the ditch. The meagre indications
preserved in casual accounts confirm this. There was an inner castle
of about 2 acres, just the area of the present garden enclosure, and
an outer court, probably an addition, of some 4 acres.[782] Both areas
were moated. Weston, a prisoner who was confined in the keep of this
castle in the 17th century, has left an account of his captivity,
in which he casually mentions that the keep or dungeon stood upon a
high terrace, from which he could overlook the outer bailey, and was
surrounded by a moat filled with water.[783]

The castle is not mentioned in Domesday, but as might be expected in a
district which had been so ravaged by war, the value of the manor had
fallen.

WORCESTER.--This borough, as we have seen, was fortified by Ethelfleda
and her husband Ethelred in the 9th century. That the fortifications
thus erected were those of a city and not of a castle is shown with
sufficient clearness by the remarkable charter of this remarkable pair,
in which they declare that they have built the _burh_ at Worcester to
shelter all the people, and the churches, and the bishop.[784] The
castle is first mentioned in the _Anglo-Saxon Chronicle_ in 1088, and
it is to be noted that it is styled the king's castle. Urse d'Abitot,
the Norman sheriff of Worcester, has the credit of having built the
first castle, and Malmesbury relates that he seized part of the monks'
cemetery for the bailey.[785] The monks, however, held on to their
right, and in the first year of Henry III. the bailey was restored to
them by the guardians of the young king, the motte being reserved for
the king's use.[786] The first wooden castle was burnt in 1113.[787]
The tower or keep which succeeded it, and which was repaired by Henry
II.,[788] may have been either of stone or wood; but in the order of
John, that the gateway of the castle, which is of wood, is to be made
of stone, we get a hint of the gradual transformation of the castle
from a wooden to a stone fortress.[789]

Worcester Castle was outside the town, from Speed's map, and was near
the Severn. The area now called College Green was no doubt the outer
ward of the castle, which was restored to the convent by Henry III.
The tower called Edgar's Tower was built by the monks as the gatehouse
to their newly conceded close.[790] From the map given by Green, this
outer bailey appears to have been roughly square; but there was also
a small oblong inner ward, retained by the king, where the gaol was
afterwards built. The area of the castle is said to have been between
3 and 4 acres.[791] The motte, which is mentioned several times in
mediæval documents,[792] was completely levelled in 1848; it was then
found out that it had been thrown up over some previous buildings,
which were believed to be Roman, though this seems doubtful.[793]

The value of Worcester had risen since the Conquest.[794]

YORK (Fig. 39).--William the Conqueror built two castles at York, and
the mottes of both these castles remain, one underneath Clifford's
Tower, the keep of York Castle, the other, on the south side of the
Ouse, still bearing the name of the Baile Hill, or the Old Baile.[795]
The _Anglo-Saxon Chronicle_ implies, though it does not directly
state, that both these castles were built in 1068, on the occasion of
William's first visit to York. The more detailed narrative of Ordericus
shows that one was built in 1068, and the other at the beginning of
1069, on William's second visit.[796] Both were destroyed in September
1069, when the English and Danes captured York, and both were rebuilt
before Christmas of the same year, when William held his triumphant
Christmas feast at York.

This speedy erection, destruction, and re-erection is enough to
prove that the castles of William in York were, like most other
Norman castles, hills of earth with buildings and stockades of wood,
especially as we find these hills of earth still remaining on the
known sites of the castles. And we may be quite sure that the Norman
masonry, which Mr Freeman pictures as so eagerly destroyed by the
English, never existed.[797] But the obstinate tendency of the human
mind to make things out older than they are has led to these earthen
hills being assigned to Britons, Romans, Saxons, Danes, anybody rather
than Normans. A single passage of William of Malmesbury, in which he
refers to the _castrum_ which the Danes had built at York in the reign
of Athelstan, is the sole vestige of basis for the theory that the
motte of Clifford's Tower is of Danish origin.[798] The other theories
have absolutely no foundation but conjecture. If Malmesbury was quoting
from some older source which is now lost, it is extremely probable
that the word _castrum_ which he copied, did not mean a castle in our
sense of the word at all, but was a translation of the word _burh_,
which almost certainly referred to a vallum or wall constructed round
the Danish suburb outside the walls of York. Such a suburb there was,
for there in 1055 stood the Danish church of St Olave, in which Earl
Siward was buried, and the suburb was long known as the Earlsburgh
or Earl's Burh, probably because it contained the residence of the
Danish earls of Northumbria.[799] This suburb was not anywhere near
Clifford's Tower, but in quite a different part of the city. To prove
that both the mottes were on entirely new sites, we have the assurance
of Domesday Book that out of the seven _shires_ or wards into which the
city was divided, one was laid waste for the castles; so that there
was clearly a great destruction of houses to make room for the new
castles.[800]

[Illustration: FIG. 39. York Castle and Baile Hill. (From a plan by P.
Chassereau, 1750.)]

What has been assumed above receives striking confirmation from
excavations made recently (1903) in the motte of Clifford's Tower.
At the depth of 13 feet were found remains of a wooden structure,
surmounted by a quantity of charred wood.[801] Now the accounts of
the destruction of the castles in 1069 do not tell us that they were
burned, but thrown down and broken to pieces.[802] But the keep
which was restored by William, and on the repair of which Henry II.
spent 15_l._ in 1172,[803] was burnt down in the frightful massacre
of the Jews at York Castle in 1190.[804] The excavations disclosed
the interesting fact that this castle stood on a lower motte than
the present one, and that when the burnt keep was replaced by a new
one the motte was raised to its present height, "an outer crust of
firmer and more clayey material being made round the older summit,
and a lighter material placed inside this crater to bring it up to
the necessary level." This restoration must have taken place in the
third year of Richard I., when 28_l._ was spent "on the work of the
castle."[805] This small sum shows that the new keep also was of wood;
and remains of timber work were in fact found on the top of the motte
during the excavations, though unfortunately they were not sufficiently
followed up to determine whether they belonged to a wooden tower or
to a platform intended to consolidate the motte.[806] It is extremely
likely that this third keep was blown down by the high wind of 1228,
when 2s. was paid "for collecting the timber of York Castle blown down
by the wind."[807] In its place arose the present keep, one of the
most remarkable achievements of the reign of Henry III.[808] The old
ground-plan of the square Norman keep was now abandoned, and replaced
by a quatrefoil. The work occupied thirteen years, from the 30th to
the 43rd Henry III., and the total sum expended was 1927_l._ 8_s._
7_d._, equal to about 40,000_l._ of our money. This remarkable fact has
slumbered in the unpublished _Pipe Rolls_ for 700 years, never having
been unearthed by any of the numerous historians of York.

The keep was probably the first work in stone at York Castle, and for
a long time it was probably the only defensive masonry. The banks
certainly had only a wooden stockade in the early part of Henry III.'s
reign, as timber from the forest of Galtres was ordered for the repair
of breaches in the _palicium_ in 1225.[809] As late as Edward II.'s
reign there was a _pelum_, or stockade, round the keep, on top of a
_murus_, which was undoubtedly an earthen bank.[810] At present the
keep occupies the whole top of the motte except a small _chemin de
ronde_, but the fact so frequently alluded to in the writs, that a
stockade ran round the keep, proves that a small courtyard existed
there formerly, as was usually the case with important keeps. Another
writ of Edward II.'s reign shows that the motte was liable to injury
from the floods of the River Fosse,[811] and probably its size has thus
been reduced.

The present bailey of York Castle does not follow the lines of the
original one, but is an enlargement made in 1825. A plan made in 1750,
and reproduced here, shows that the motte was surrounded by its own
ditch, which is now filled up, and that the bailey, around which a
branch of the Fosse was carried, was of the very common bean-shaped
form; it was about 3 acres in extent. The motte and bailey were both
considerably outside what is believed to have been the Anglo-Saxon
rampart of York,[812] but the motte was so placed as to overlook the
city.

The value of the city of York, in spite of the sieges and sacks which
it had undergone, and in spite of there being 540 houses "so empty
that they pay nothing at all," had risen at the date of the Survey
from 53_l_. in King Edward's time to 100_l._ in King William's.[813]
This extraordinary rise in value can only be attributed to increased
trade and increased exactions, the former being promoted by the greater
security given to the roads by the castles, the latter due to the tolls
on the high-roads and waterways, which belonged to the king, and the
various "customs" belonging to the castles, which, though new, were
henceforth equally part of his rights.

THE BAILE HILL, York (Fig. 39).--There can be no doubt whatever that
this still existing motte was the site of one of William's castles
at York, and it is even probable that it was the older of the two,
as Mr Cooper conjectures from its position on the south side of the
river.[814] The castle bore the name of the Old Baile at least as early
as the 14th century, perhaps even in the 12th.[815] In 1326 a dispute
arose between the citizens of York and Archbishop William de Melton as
to which of them ought to repair the wall around the Old Baile. The
mayor alleged that the district was under the express jurisdiction of
the archbishop, exempt from that of the city; the archbishop pleaded
that it stood within the ditches of the city.[816] The meaning of
this dispute can only be understood in the light of facts which have
recently been unearthed by the industry and observation of Mr T. P.
Cooper, of York.[817] The Old Baile, like so many of William's castles,
originally stood outside the ramparts of the city. The original Roman
walls of York (it is believed) enclosed only a small space on the
eastern shore of the Ouse, and before the Norman Conquest the city
had far outgrown these bounds, and therefore had been enlarged in
Anglo-Saxon times. It appears that the Micklegate suburb was then for
the first time enclosed with a wall, and as this district is spoken of
in Domesday Book as "the shire of the archbishop," it was evidently
under his jurisdiction. At a later period this wall was buried in an
earthen bank, which probably carried a palisade on top, until the
palisade was replaced by stone walls in the reign of Henry III.[818]

The evidence of the actual remains renders it more than probable that
this rampart turned towards the river at a point 500 feet short of
its present angle, so that the Old Baile, when first built, was quite
outside the city walls.[819] This is exactly how we should expect to
find a castle of William the Norman's in relation to one of the most
turbulent cities of the realm; and, as we have seen, the other castle
at York was similarly placed. By the time of Archbishop Melton the
south-western suburb was already enclosed in the new stone walls built
in the 13th century, and these walls had been carried along the west
and south banks of the Old Baile, so as to enclose that castle within
the city. This was the archbishop's pretext for trying to lay upon
the citizens the duty of maintaining the Old Baile. But probably on
account of his ancient authority in this part of the city, the cause
went against him; though he stipulated that whatever he did in the way
of fortification was of his own option, and was not to be accounted
a precedent. A contemporary chronicler says that he enclosed the Old
Baile first with stout planks 18 feet long, afterwards with a stone
wall:[820] an interesting proof that wooden fortifications were still
used in the reign of Edward III.

Though the base court of the Old Baile is now built over, its area and
ditches were visible in Leland's time,[821] and can still be guessed
at by the indications Mr Cooper has noted. The area of the bailey must
have been nearly 3 acres, and its shape nearly square. This measurement
includes the motte, which was placed in the south-west corner on
the line of the banks; it thus overlooked the river as well as the
city.[822]




CHAPTER VIII

MOTTE-CASTLES IN NORTH WALES


Motte-castles are as common in Wales as they are in England, and in
certain districts much more common. It is now our task to show how they
got there. They were certainly not built (in the first instance at any
rate) by the native inhabitants, for they do not correspond to what we
know to have been the state of society in Wales during the Anglo-Saxon
period.[823] The Welsh were then in the tribal condition, a condition,
as we have shown, inconsistent with the existence of the private
castle. The residence of the king or chieftain, as we know from the
Welsh Laws, was a great hall, such as seems to have been the type of
chieftains' residence among all the northern nations at that time. "It
was adapted for the joint occupation of a number of tribesmen living
together."[824]

Pennant describes the residence of Ednowen, a Welsh chieftain of the
12th century, as follows: "The remains are about 30 yards square; the
entrance about 7 feet wide, with a large upright stone on each side
for a doorcase; the walls were formed of large stones uncemented by
any mortar; in short the structure shows the very low state of Welsh
architecture at this time; it may be paralleled only by the artless
fabric of a cattle-house."[825] This certainly is a hall and not a
castle.

The so-called Dimetian Code indeed tells us that the king is to have
a man and a horse from every hamlet, with hatchets for constructing
his castles (gestyll) at the king's cost; but the Venedotian Code,
which is the older MS., says that these hatchet-men are to form
encampments (uuesten); that is, they are to cut down trees and form
either stockades on banks or rude _zerebas_ for the protection of
the host.[826] It is clearly laid down in the Codes what buildings
the king's villeins are to erect for him at his residences: a hall,
buttery, kitchen, dormitory, stable, dog-house, and little house.[827]
In none of these lists is anything mentioned which has the smallest
resemblance to a castle, not even a tower. We can imagine that these
buildings were enclosed in an earthwork or stockade, but it is not
mentioned.[828]

Wales was never one state, except for very short periods. Normally
it was divided into three states, Gwynedd or North Wales, Powys or
Mid-Wales, and Deheubarth, all almost incessantly at war with each
other.[829] Other subdivisions asserted themselves as opportunity
offered, so that the above rough division into provinces must not be
regarded as always accurate. A Wales thus divided, and perpetually
rent by internal conflicts, invited the aggression of the Saxons, and
it is probable that the complete subjugation of Britain would have
been accomplished by the descendants of Alfred, if it had not been
for the Danish invasions. The position of the Welsh kings after the
time of Athelstan seems to have been that of tributaries, who threw
off their allegiance whenever it was possible to do so. But still
the Anglo-Saxon frontier continued to advance. Professor Lloyd has
shown, from a careful examination of Domesday Book, that even before
the Norman Conquest the English held the greater part of what is now
Flintshire and East Denbighshire, and were advancing into the vale of
Montgomery and the Radnor district.[830] The victories of Griffith ap
Llywelyn, an able prince who succeeded in bringing all Wales under
his sway, devastated these English colonies; but his defeat by Earl
Harold in 1063 restored the English ascendancy over these regions.
The unimpeachable evidence of Domesday Book shows that a considerable
district in North Wales and a portion of Radnor were held respectively
by Earl Edwin and Earl Harold before the Norman Conquest. Moreover, the
fact mentioned by the _Anglo-Saxon Chronicle_ in 1065, that Harold was
building a hunting-seat for King Edward at Portskewet, _after he had
subdued it_, suggests that the land between Wye and Usk, which Domesday
Book reckons under Gloucestershire, was a conquest of Harold's.[831]

The Norman Conqueror was not the man to slacken his hold on any
territory which had been won by the Saxons. But there is no succinct
history of his conquests in Wales; we have to make it out, in most
cases, from notices that are scarcely more than allusive, and from
the surer, though scanty, ground of documents. It is noteworthy that
the _Anglo-Saxon Chronicle_ is so hostile to the Norman kings that it
discounts their successes in Wales. Thus we have only the briefest
notice of William I.'s invasion of South Wales, which was very probably
the beginning of the conquest of that region; and several expeditions
of William II. are spoken of as entirely futile, though as we are
told that the existing castles were still held by the Normans, or new
ones were built, it is clear that this summing-up is not strictly
correct.[832] Our Welsh authorities, the _Annales Cambriæ_ and the
_Brut y Tywysogion_,[833] seem to give a fairly candid account of the
period, although the dates in the _Brut_ are for the most part wrong
(sometimes by three years), and they hardly ever give us a view of the
situation as a whole. They tell us when the Welsh rushed down and burnt
the castles built by the Normans in the conquered districts, but do
not always tell when the Normans recovered and rebuilt them.

Fortunately we are not called upon here to trace the history of the
cruel and barbarous warfare between Normans and Welsh. No one can turn
that bloodstained page without wishing that the final conquest had come
two hundred years earlier, to put an end to the tragedy of suffering
which must have been so largely the portion of the dwellers in Wales
and the Marches after the coming of the Normans.[834] Our business
with both Welsh and Normans is purely archæological. We hold no brief
for the Normans, nor does it matter to us whether they kept their
hold on Wales or were driven out by the Welsh; our concern is with
facts, and the solid facts with which we have to deal are the castles
whose remains still exist in Wales, and whose significance we have to
interpret.

"Wales was under his sway, and he built castles therein," says the
_Anglo-Saxon Chronicle_, in summing up the reign of the Conqueror; a
passage which is scarcely consistent with its previous almost complete
silence about events in Wales. There can be little doubt that William
aimed at a complete conquest of Wales, and that the policy he adopted
was the creation of great earldoms along the Welsh border, endowed with
special privileges, one of which was the right of conquering whatever
they could from the Welsh.[835] To these earldoms he appointed some
of his strongest men, men little troubled by scruples of justice or
mercy, but capable leaders in war or diplomacy. It was an essential
part of the plan that every conquest should be secured by the building
of castles, just as had been done in England. And we have now to trace
very briefly the outline of Norman conquest in Wales by the castles
which they have left behind them.

We shall confine ourselves to those castles which are mentioned in the
_Brut y Tywysogion_, the _Pipe Rolls_, or other trustworthy documents
between 1066 and 1216, the end of King John's reign. Of many of these
castles only the earthworks remain; of many others the original plan,
exactly similar to that of the early castles of Normandy and France, is
still to be traced, though masked by the masonry of a later age. Grose
remarked but could not explain the fact that we continually read of the
castles of the Marches being burnt and utterly destroyed, and a few
months later we find them again standing and in working order. This can
only, but easily, be explained when we understand that they were wooden
castles built on mottes, quickly restored after a complete destruction
of the wooden buildings.

North Wales appears to have been the earliest conquest of the Normans,
though not the most lasting. North Wales comprised the Welsh kingdoms
of Gwynedd and Powys. Gwynedd covered the present shires of Anglesea,
Carnarvon, and Merioneth, and the mountainous districts round
Snowdon.[836] Powys stretched from the estuary of the Dee to the upper
course of the Wye, and roughly included Flint, Denbigh, Montgomery,
and Radnor shires. Hugh of Avranches, Earl of Chester, was the great
instrument of Norman conquest in Gwynedd, and in the northern part
of Powys, which lay so near his own dominions. He was evidently a man
in whose ability William had great confidence, as he removed him from
Tutbury to the more difficult and dangerous position of Chester, and
gave his earldom palatine privileges; all the land in Cheshire was held
under the earl, and he was a sort of little king in his county.

Hugh appears to have at once commenced the conquest of North Wales. As
Professor Lloyd remarks, Domesday Book shows us Deganwy as the most
advanced Norman post on the North Welsh coast, while on the Bristol
Channel they had got no further than Caerleon.[837] In advancing to the
valley of the Clwyd and building a castle at Rhuddlan, the Normans were
only securing the district which had already been conquered by Harold
in 1063, when he burnt the hall of King Griffith at Rhuddlan. Nearly
the whole of Flintshire (its manors are enumerated by Domesday Book
under Cheshire) was held by Earl Hugh in 1086, so that he commanded
the entire road from Chester to Rhuddlan. His powerful vassal, Robert
of Rhuddlan, who became the terror of North Wales, besides the lands
which he held of Earl Hugh, held also directly of the King Rhos and
Rhufeniog, districts which roughly correspond to the modern shire of
Denbigh, and "Nort Wales" which Professor Lloyd takes to mean the
remainder of the principality of Gwynedd, from which the rightful
ruler, Griffith ap Cynan, had been driven as an exile to Ireland.

It does not appear that there was any fortification at RHUDDLAN[838]
before the "castle newly erected" by Earl Hugh and his vassal Robert.
They shared between them the castle and the _new borough_ which was
built near it.[839] One word about this new borough, which will apply
to the other boroughs planted by Norman castles. There were no towns in
Wales of any importance before the Norman Conquest, and this civilising
institution of the borough is the one great set-off to the cruelty and
unrighteousness of the conquest. Mills, markets, and trade arose where
castles were seated, and civilisation followed in their train.

The castle of Hugh and Robert was not the magnificent building which
still stands at Rhuddlan, for that is entirely the work of Edward I.,
and there is documentary evidence that Edward made a purchase of new
land for the site of his castle.[840] More probably Robert and Hugh had
a wooden castle on the now reduced motte which may be seen to the south
of Edward's castle. In Gough's time this motte was still "surrounded
with a very deep ditch, including the abbey." Nothing can be seen of
this ditch now, except on the south side of the motte, where a deep
ravine runs up from the river. As from Gough's description the hillock
(called Tut Hill)[841] was within the precincts of the priory of Black
Friars, founded in the 13th century, it is extremely probable that
Edward gave the site of the old castle to the Dominicans when he built
his new one.[842]

Another of the castles of Robert of Rhuddlan was DEGANWY, or Gannoc,
which defended the mouth of the Conway.[843] Here it is said that there
was an ancient seat of the kings of Gwynedd.[844] The two conical hills
which rise here offer an excellent site for fortification, one of them
being large enough on top for a considerable camp. The Norman Conqueror
treated them as two mottes, and connected them by walls so as to form
a bailey below them. The stone fortifications are probably the remains
of the castle built by the Earl of Chester in 1211.[845] This castle
was naturally a sorely contested point, and often passed from hand to
hand; but it was in English possession in the reign of Henry III. It
was abandoned when Edward I. built his great castle at Conway.

With its usual indifference, the Survey mentions no castle in
Flintshire, but we may be sure that the castle of MOLD, or Montalto
(Fig. 40), was one of the earliest by which the Norman acquisitions in
that region were defended,[846] though it is not mentioned in authentic
history until 1147. The tradition that it was built by Robert de Monte
Alto, one of the barons of the Earl of Chester, is no doubt correct,
though the assumption of Welsh legend-makers that the _Gwydd Grug_, or
great tumulus, from which this castle derives its Welsh name, existed
before the castle, may be dismissed as baseless. The motte of Robert de
Monte Alto still exists, and is uncommonly high and perfect; it has two
baileys, separated by great ditches, and appears to have had a shell on
top. [D. H. M.] The castle was regarded as specially strong, and its
reduction by Owen Gwynedd in 1147 was one of the sweetest triumphs that
the Welsh ever won.[847]

[Illustration: FIG. 40.--MOTTE-CASTLES OF NORTH WALES. Mold. Welshpool.
Wrexham. Mathraval.]

It is clear from the _Life of Griffith ap Cynan_[848] that the Earl of
Chester had conquered and incastellated Gwynedd before the accession
of William Rufus. This valuable document unfortunately gives no dates,
but it mentions in particular the castle at Aberlleinog,[849] one
at Carnarvon, one at Bangor, and one in Merioneth. The motte at
ABERLLEINOG, near Beaumaris, still exists, and the half-moon bailey
is traceable, but the curious little round towers and revetting wall
in masonry on the motte were probably built to carry guns at the time
of the Civil War, when this castle was besieged by the Royalists.
At CARNARVON the magnificent castle of Edward I. has displaced all
former erections, yet some evidence for a motte-and-bailey plan may
be found in the fact that the northern portion of the castle has
evidently been once separated by a ditch from the southern, and is
also much higher.[850] On the hills above BANGOR, Pennant thought he
had discovered the remains of Earl Hugh's castle, but having carefully
examined these walls, we are convinced that they never formed part of
a castle at all, as they are much too thin; nor are there any vestiges
of earthworks.[851] We are disposed to think that instead of at Bangor,
the castle of Earl Hugh was at ABER, often spoken of as ABERMENAI in
the _Chronicles_, and evidently the most important port on the Straits.
At Aber there still remains a motte which must have belonged to an
important castle, as it was afterwards one of the seats of Llywelyn
ap Jorwerth, Prince of Gwynedd. The castle in Merioneth cannot be
certainly identified.

In one of the invasions of William Rufus, which both the _Anglo-Saxon
Chronicle_ and the _Brut_ describe as so unsuccessful, we hear that he
encamped at MUR CASTELL, a place undoubtedly the same as what is now
called TOMEN-Y-MUR, a motte standing just inside a Roman camp, on the
Roman road leading from Shropshire into Merioneth and Carnarvon. This
motte is surrounded by a ditch; there are traces of the usual earthen
rampart round the top, now mutilated by landslips.[852] We may, with
great probability, assume that this motte was thrown up by William
Rufus, and that the Roman camp served as a bailey for his invading
host. Whether it was garrisoned for the Normans we cannot say, but it
evidently formed an important post on a route often followed by their
invading armies, as Henry I. is said to have encamped there twice.[853]
It is one of the few mottes which stand in a wild and mountainous
situation, and its purpose no doubt was purely military.[854]

       *       *       *       *       *

The earls of Chester did not retain the sovereignty of Gwynedd; on the
death of Rufus, Griffith ap Cynan returned, and obtained possession
of Anglesea. He was favourably received at the court of Henry I., and
gradually recovered possession of the whole of Gwynedd. In 1114 Henry
had to undertake a great expedition against him to enforce the payment
of tribute;[855] from which, and from the peaceful manner in which
Griffith seems to have acquired his principality, we may infer that
this tribute was the bargain of his possession. It very likely suited
Henry's policy better to have a tributary Welsh prince than a too
powerful earl of Chester.

The reigns of the three first Norman kings were the time in which
Norman supremacy in Wales made its greatest advances. With the
accession of Stephen and the civil war which followed it came the
great opportunity for the Welsh of throwing off the Norman yoke. Powys
appears to have been the only province which remained faithful to the
English allegiance, under Madoc ap Meredith.[856] The history of Norman
conquest in Powys is more confused than that of Gwynedd, but Domesday
shows us that Rainald, the Sheriff of Shropshire, a vassal of Earl
Roger of Shrewsbury, was seated at Edeyrnion and Cynlle, two districts
along the upper valley of the Dee.[857] Robert of Rhuddlan held part of
his grant of "Nort Wales," namely the hundred of Arwystli, in the very
centre of Wales, under Earl Roger. Professor Lloyd remarks, "Earl Roger
claimed the same authority over Powys as Earl Hugh over Gwynedd, and
the theory that the princes of this region were subject to the lords of
Salop survived the fall of the House of Montgomery."[858]

We have already spoken of Earl Roger de Montgomeri and his brood of
able and unscrupulous sons.[859] The palatine earldom of Shrewsbury
lay along the eastern border of central Powys, and must soon have
proved a menace to that Welsh kingdom. Domesday Book shows us that
Earl Roger had already planted his castle of Montgomery well within
the Welsh border at that date. But the ambition of Earl Roger and his
sons stretched beyond their immediate borders. It is probable that
they used the upper Severn valley, which they fortified by the castle
of Montgomery, and possibly by the castle of Welshpool, as their road
into Ceredigion, for we find Earl Roger named by the _Brut_ as the
builder of the castle of Cilgerran,[860] and some say of Cardigan also.
Possibly he was helping his son Arnolf in the conquest of Pembroke. In
1098 we find his successor, Earl Hugh, allied with the Earl of Chester
in the invasion of Anglesea.

MONTGOMERY.--This castle is named from the ancestral seat of its
founder.[861] The motte-and-bailey plan is still very apparent in the
ruins, though the motte is represented by a precipitous rock, only a
few feet higher than the baileys attached, and separated from them
by a ditch cut through the headland. The masonry, the chief part of
which is the shell wall and towers on this isolated rock, is none of
it older than the reign of Henry III., when large sums were spent
on this castle, and it is spoken of in a writ as "the new Castle of
Montgomery."[862] Yet even then the whole of the defences were not
remade in stone, as bretasches of timber are ordered in a _mandamus_ of
1223.[863] The four wards are all roughly rectilateral. The castle was
never recovered permanently by the Welsh, and after the forfeiture of
Robert Belesme, the third Earl of Shrewsbury, in 1101, the Crown kept
this important border fortress in its own hands throughout the Middle
Ages.

Although Montgomery Castle is the only one mentioned in that region
at the same date, there must have been many others, for in 1225 Henry
III. ordered all who had mottes in the valley of Montgomery to fortify
them with good bretasches without delay;[864] and the remains of these
mottes are still numerous in the valley. It is quite possible that the
mottes at Moat Lane and Llandinam were thrown up to defend the road
into Arwystli; but this is conjecture.[865]

WELSHPOOL, _alias_ Pol or Pool (Fig. 40), is also called the Castle of
Trallung.--In Powell's _History of Wales_ (p. 137) it is stated that
Cadwgan ap Bleddyn, when Henry I. took Cardigan from him, retired to
Powys, and began to build a castle here. Powell's statements, however,
have no authority when unconfirmed, and we are unable to find any
confirmation of this statement in the more trustworthy version of the
_Brut_. And as the House of Montgomeri was firmly established in the
valley of Montgomery as early as 1086, it seems more probable that
the two motte-and-bailey castles at Welshpool, lower down the Severn
valley, are relics of the early progress of that family, especially
as one of these castles is only about a mile east of Offa's Dyke, the
ancient border. This latter motte is partly cut into by the railway,
and diminished in size, but the bailey is nearly perfect. The other
one is in the park of Powys Castle, and is an admirable specimen of
its class. The breastwork round the top of the motte remains. [H. W.]
It seems probable that this was the precursor of Powys Castle, and was
abandoned at an early period, as the newer castle was known by the name
of Castell Coch, or the Red Castle, as early as 1233.[866] Leland
states that there were formerly two castles of two different Lords
Marchers at Welshpool;[867] possibly this throws some light on the
existence of these two motte-castles.

       *       *       *       *       *

When Henry II. came to the throne in 1154, one of the many questions
which he had to settle was the Welsh question. His first expedition
against North Wales was in 1157. Here he was one day placed in grave
difficulties, and fortune was only restored by his personal courage.
But in spite of this we learn even from the Welsh chronicler that
he continued his advance to Rhuddlan, and that the object of the
expedition, which was the restoration of Cadwalader, one of the sons
of Griffith ap Cynan, to his lands, was accomplished. The English
chronicler Roger of Wendover says that Henry recovered all the
fortresses which had been taken from his predecessors, and rebuilt
Basingwerk Castle; and when he had reduced the Welsh to submission,
returned in triumph to England. The undoubted facts of the _Pipe Rolls_
show us that in the year 1159 Henry had in his hands the castles of
Overton, Hodesley, Wrexham, Dernio, Ruthin, and Rhuddlan, castles
which would give him command of the whole of Flintshire and of East
Denbigh and the valley of the Clwyd. Similarly, after the expedition
of 1165, sometimes stated to have been only disastrous, we find him in
possession of the castles of Rhuddlan, Basingwerk, Prestatyn, Mold,
Overton, and Chirk;[868] so that after the battle of Crogen, or Chirk,
he actually held the battlefield.

We are thus introduced to an entirely new group of castles, Rhuddlan
being the only one which we have heard of before. But it is highly
probable that most of these castles were originally raised by the earls
of Chester or Shrewsbury, and were in Henry's hands by escheat.

*BASINGWERK.--The _werk_ referred to in this name has probably nothing
to do with the castle, but refers to Wat's Dyke, which reaches the
Dee at this point. The abbey at this place was founded by an earl
of Chester,[869] which makes it probable that the castle also was
originally his work, especially as Wendover says that Henry _rebuilt_
it. There is no trace of a castle near the abbey,[870] but less than
a mile off, near Holywell Church, there is a headland called Bryn
y Castell, with a small mound at the farther end, which has far
more claim to be the site of Basingwerk Castle, especially as it is
mentioned in John's reign (when it was retaken from the Welsh) as the
castle of Haliwell.[871]

OVERTON, in East Denbigh, on the middle course of the Dee. In custody
of Roger de Powys for the king in 1159-1160. As Leland speaks of the
ditches and hill of the castle, it was probably a motte-castle of the
usual type. "One parte of the ditches and Hille of the castel yet
remaynith; the residew is in the botom of Dee."[872] It is probably all
there now, as not a vestige can be traced. [B. T. S.]

DERNIO, or Dernant.--There can be no question that Dernio is
Edeyrnion, the valley stretching from Bala Lake to Corwen. Domesday
Book tells us that Rainald the Sheriff, a "man" of Earl Roger of
Shrewsbury, held two "fines" in Wales, Chenlei and Dernio, that is,
Cynllaith and Edeyrnion.[873] Towards the end of the 11th century there
must have been a Norman castle at Rug in Edeyrnion, as it was to this
place that the earls of Chester and Shrewsbury enticed Griffith ap
Cynan, the rightful ruler of Gwynedd; they then sent him prisoner to
Chester for twelve years.[874] Very likely the castle of Dernio, which
Henry II. was putting into a state of defence in 1159,[875] was at RUG,
1-1/2 miles from Corwen, where there is still a motte in some private
grounds, and there was formerly a bailey also.[876] The place was the
seat of an important family in later times. At any rate, the castle was
in Edeyrnion, and shows that Henry was holding the northern part of
Merionethshire.

HODESLEY; undoubtedly "The Rofts" near Gresford, a motte with remains
of a bailey, on a headland above the river Alyn. It is in the former
lordship of Hoseley.[877]

WREXHAM, the Wristlesham of the _Pipe Rolls_ (Fig. 40).--Henry was
paying for the custody of this castle and that of Hoseley in 1160 and
1161. Both castles are in the district of Bromfield, which was one of
the early acquisitions of the earls of Chester. Mr Palmer remarks that
this district was probably ceded to the princes of Powys, in return for
the help which they often rendered to the English king against other
Welsh princes, as it is found as part of Powys at a later period.[878]
There are no remains of any castle at Wrexham itself, but about a mile
off, in Erddig Park, there is a motte and bailey of considerable size
(though the motte is reduced) showing that a castle of some importance
once stood there. There were formerly some remains of masonry.[879]
Wat's Dyke has been utilised to form one side of the bailey. It is
probable that the importance of the two Bromfield castles, Wrexham and
Hoseley, was lost when the princes of Powys built their castle on Dinas
Bran.

*RUTHIN.--This important castle, defending the upper valley of the
Clwyd, was probably in existence long before Henry II. repaired it in
1160, and may perhaps be attributed to Earl Hugh of Chester. The plan
shows distinctly that it was once a motte and bailey, though the castle
is now transformed into a modern house.[880]

CHIRK, or Crogen, in the valley of the Ceiriog.--Henry was paying
for the custody of this castle in 1164, and was provisioning it in
1167.[881] King John paid for the erection of a bretasche there,
possibly after some destruction by the Welsh.[882] Probably the first
castle of Chirk did not stand in the commanding situation now occupied
by the castle of Edward I.'s reign, but is represented by a small
motte in a garden near the Ceiriog stream, and close to the church. An
Anglo-Norman poem of the 13th century attributes the first building of
this castle to William Peverel, Lord of Whittington and Ellesmere, and
says he placed it "on the water of Ceiriog."[883] No doubt it defended
the passage of the stream, and an important road into Shropshire.

PRESTATYN.--This castle defended the coast road from Chester to
Rhuddlan. Henry II. granted it to Robert Banaster for his services
in 1165.[884] It was destroyed by Owen Gwynedd in 1167, and does not
appear to have been rebuilt. A low motte with a half-moon bailey, and a
larger square enclosure, still remain. [B. T. S.]

       *       *       *       *       *

Mr Davis has remarked that John was more successful in extending his
authority over the British Isles than in anything else.[885] In 1211 he
led an expedition into the heart of Wales, and reduced his son-in-law
Llywelyn ap Jorwerth to complete submission. As usual, the expedition
was marked by the building or repair of castles. The Earl of Chester
restored Deganwy, which shows that the English frontier was again
advanced to the Conway; he also repaired the castle of Holywell, which
the _Pipe Roll_ shows to have been recovered from the Welsh about this
time.[886] These _Rolls_ also show that in 1212-1213 John was paying
for works at the castles of Carreghova, Ruthin, and Chirk, as well as
at the following castles, which have not been mentioned before.

MATHRAVAL, Madrael in the _Pipe Rolls_ (Fig. 40), near Meifod in
Montgomeryshire, defending the valley of the Vyrnwy.--Here was the
chief royal residence of Powys;[887] but the castle was built in John's
reign by Roger de Vipont. It occupied 2-1/4 acres, and the motte is
in one corner of the area, which is square,[888] and surrounded only
by banks; though ruined foundations are found in parts of the castle.
John himself burned the castle in 1211, when the Welsh were besieging
it,[889] but the _Pipe Roll_ (1212-1213) shows that he afterwards
repaired it. [D. H. M.]

EGLOE, or Eulo, called by Leland Castle Yollo.--On the Chester and
Holywell road, about 8 miles from Holywell. The mention in the _Pipe
Roll_ of pikes and ammunition provided for this castle in 1212-1213 is
the first ancient allusion to it with which we are acquainted. It is a
motte-and-bailey castle, with additions in masonry which are probably
of the reign of Henry III. The keep is of the "thimble" plan, a rare
instance.[890] [B. T. S.]

*YALE.--The _Brut_ tells us that in 1148 (read 1150) Owen Gwynedd built
a castle in Yale. Powell identified this with Tomen y Rhodwydd, a motte
and bailey on the road between Llangollen and Ruthin. Yale, however, is
the name of a district, and there can be little doubt that the castle
of Yale was the motte and bailey at Llanarmon, which for a long period
was the _caput_ of Yale.[891] Yale undoubtedly belonged to the Normans
when Domesday Book was compiled,[892] and it is therefore not unlikely
that these earthworks were first thrown up by the Earl of Chester. The
castle was burnt by Jorwerth Goch in 1158, but restored by John in
1212. One of the expenses entered for that year is "for iron mallets
for breaking the rocks in the ditch of the castle of Yale."[893] This
ditch cut in the rock still remains, as well as some foundations on
the motte,[894] which is known as Tomen y Vardra, or the Mount of the
demesne.[895]

       *       *       *       *       *

How long the two last-mentioned groups of castles continued in
Anglo-Norman hands we do not attempt to say. North Wales, as is well
known, reaped a harvest of new power and prosperity through the
civil war of the end of John's reign, and the ability of Llywelyn ap
Jorwerth. Our task ends with the reign of John. We have only to remark
that until the _Pipe Rolls_ of Henry III.'s reign have been carefully
searched, it is impossible to say with certainty what castles of North
Wales, or if any, were still held by the English king.




CHAPTER IX

MOTTE-CASTLES IN SOUTH WALES


It is not possible to fix certain dates for all the Norman conquests of
the several provinces of South Wales. These conquests proceeded from
various points, under different leaders. We might have expected that
the earliest advances would have been on the Herefordshire border,
the earldom of Hereford having been given by William I. to William
FitzOsbern, one of his most trusted and energetic servants. Ordericus
tells us that FitzOsbern and Walter de Lacy first invaded the district
of Brecknock, and defeated three kings of the Welsh.[896] This looks
as though the conquest of Brecknock was then begun. But it was not
completed till the reign of Rufus; in 1093 Bernard of Neufmarché
defeated and slew Rhys ap Tudor, King of South Wales, in a battle
which the Welsh chronicler speaks of as the fall of the kingdom of the
Britons.[897] William FitzOsbern died in 1071, and he had scarcely time
to accomplish more than the building of the border castles of Wigmore,
Clifford, Ewias, and Monmouth, and the incastellation of Gwent, that
is the country between the Wye and the Usk, which had already been
conquered by Harold.

It seems probable that Pembrokeshire was one of the earliest Norman
conquests in South Wales, as in 1073 and 1074 the _Brut_ tells of
two expeditions of "the French" into Dyfed, a region which included
not only what we now call Pembrokeshire, but also Strath Towy, which
comprised an extensive district on both sides of the valley of the
Towy.[898] The _Annales Cambriæ_ name Hugh de Montgomeri, Earl Roger's
eldest son, in connection with the second of these expeditions, seven
years before the expedition of King William into Wales in 1081.[899]
The House of Montgomeri certainly took the most conspicuous part in
the conquest of Dyfed and Cardigan, which was completed, according
to the _Brut_, in 1093.[900] Arnulf of Montgomeri, fifth son of Earl
Roger, was the leader of this conquest. But his father must at the same
time have been operating in Cardigan, as the building of the castle of
Cilgerran, which is on the very borders of Pembroke and Cardigan, is
attributed to him.

How far Earl Roger made himself master of Ceredigion it is impossible
to say. Later writers say that he built the castle of Cardigan, but
we have not been able to find any early authority for this statement,
which in itself is not improbable. Powell's _History_ makes him do
homage to William Rufus for the lordship of Cardigan, but here again
the authority is doubtful.[901] The fact that a castle in or near
Aberystwyth was not built until 1109 may indicate that the conquest
of Northern Cardigan was not completed till it became the portion
of the De Clares. This took place in 1109, when Henry I. deposed
Cadwgan, a Welsh prince whom he had made Lord of Cardigan, and gave
the lordship to Gilbert de Clare, who immediately proceeded to build
the above-mentioned castle, and to restore Earl Roger's castle at
Cilgerran (Dingeraint).[902] From this time the castle and district of
Cardigan continued to be an appanage of the House of Clare (of course
with frequent interruptions from Welsh invasions), and of the family
of William Marshall, to whom the Clare lands came by marriage. The
authority of these earls was suspended during the reign of Henry II.,
when he made Rhys ap Griffith, who had possessed himself of Ceredigion
by conquest, Justiciar of South Wales, but in the reigns of John and
Henry III., the _Close Rolls_ show that Cardigan Castle and county were
generally in the hands of the Marshalls.

The conquest of Pembrokeshire must have been closely followed by that
of what is now Carmarthenshire, which was then reckoned as part of
Dyfed.[903] We first hear of the castle of Rhyd y Gors in 1094,[904]
but it evidently existed earlier. This castle we believe to have been
the important castle of Carmarthen (see _post_). It was founded by
William, son of Baldwin, sheriff of Devon, and cousin of the Gilbert
de Clare who at a later period was made Lord of Cardigan by Henry
I. We thus see at what an early date this important family made its
appearance in Welsh history.

The conquest of Brecknock (Brecheiniog) we have already briefly
referred to. It must have begun as early as 1088, for in that year
Bernard de Neufmarché gave to St Peter's Abbey at Gloucester the church
and manor of Glasbury. The inheritance of Bernard passed by marriage
to the De Braoses, and from them to the Mortimers. It is convenient
to mention in this connection the Norman conquest of Radnor, of which
the De Braoses and Mortimers were the heroes. A charter of Philip
de Braose, not later than 1096, is dated at "Raddenoam."[905] Even
during the anarchy of Stephen's reign, the Mortimers were able to
maintain their hold on this district, for the _Brut_ relates that in
1145, Hugh, son of Ralph Mortimer, conquered Malienydd and Elvael the
second time.[906] These two districts properly belong to Powys, though
geographically in South Wales.

We leave to the last the conquest of Glamorgan, which may possibly have
been one of the earliest, but whose date is still a matter of dispute,
owing to the legendary nature of the Aberpergwm version of the _Brut_,
the only one which even alludes to this conquest. We have, however, an
initial date given us in the year 1082, when the _Brut y Tywysogion_
tell us of the building of Cardiff Castle.[907] The conquest of
"Morgannwg," that is the country between the Usk and the Neath, was the
most permanent of any of those accomplished by the Normans in Wales,
but its details are the most obscure of any. The earlier version of
the _Brut_ takes no notice of the conquest of Glamorgan; the later
version which goes by the name of the _Gwentian Chronicle_[908] tells
us that the Norman Robert Fitz Hamon, being called in to the help of
one Welsh prince against another, conquered Glamorgan for himself,
and divided it amongst his followers, who built castles in all parts
of the country. The date given is 1088. It seems to be agreed by
historians that while the facts of Robert Fitz Hamon's existence and
of his conquest of Glamorgan are certain, the details and the list of
followers given in this chronicle are quite untrustworthy.[909]

The district called Gower did not then form part of Glamorgan, as it
does now, though it is still ecclesiastically separate. If we are to
believe the Aberpergwm _Brut_, it must have been conquered in 1094,
when William de Londres, one of the "knights" of Robert Fitz Hamon,
built a strong castle in Cydweli (Kidwelly).[910]

We will now briefly notice such of the castles of these various
districts as are mentioned in the sources to which we have already
referred in our last chapter, taking them in the order of the modern
counties in which they are found.


CASTLES OF PEMBROKESHIRE.

PEMBROKE.--Giraldus says that Arnulf de Montgomeri first built this
castle of sods and wattles, a scanty and slender construction, in the
reign of Henry I.[911] This date, however, must certainly be wrong,
for the castle sustained a siege from the Welsh in 1094, and in 1098
Arnulf gave the chapel of St Nicholas in his castle of Pembroke to the
abbey of St Martin at Sées.[912] There is no motte at Pembroke Castle;
the magnificent keep (clearly of the 13th century or later) stands
in a small ward at the edge of a cliff,[913] separated by a former
ditch from the immense encircling bailey whose walls and towers are
clearly of Edwardian date. The words of Giraldus "a castle of wattles
and turf" might lead us to think that the first castle was a motte of
the usual type, but the use which he makes of the same expression in
his work on Ireland leads one to think that he means a less defensible
fort, a mere bank and fence.[914] There is some reason, moreover, to
doubt whether the present castle of Pembroke stands on the same site
as Arnulf's, as after the banishment of the latter, Gerald, the royal
Seneschal of Pembroke "built the castle anew in the place called Little
Cengarth."[915]

But however this may be, the castle of Pembroke was certainly strong
enough in 1094 to resist a great insurrection of the Welsh, when all
the castles of south-west Wales were destroyed, except Pembroke and
Rhyd y Gors. And it continued to be one of the chief strongholds of
English power in South Wales until Edward I. completed the conquest of
the country. Its splendid situation on a high cliff at the mouth of an
excellent harbour, to which supplies could be brought by sea, was one
of the secrets of its strength. A passage cut in the rock led from the
castle to a cave below opening on to the water.

*NEWPORT, or Trefdaeth, was the head of the Barony of Keymes, an
independent lordship founded at the time of the first Norman advance,
by Martin of Tours.[916] There is no mention of it before 1215. The
present ruined castle of Newport is not earlier than the 13th century,
but about 1-1/2 miles higher up the river, at Llanhyfer, is a fine
motte and bailey, which probably mark the site of the first castle of
Martin of Tours.[917]

WISTON, _alias_ Gwys or Wiz.--First mentioned in 1148, when it was
taken by the Welsh.[918] At a later period we find it one of the
castles of the Earl of Pembroke. There is a motte still remaining,
with a shell wall on top, 6 feet thick, having a plain round arched
entrance. This masonry is probably the work of William Marshall, Earl
of Pembroke, as he restored the castle in 1220 after it had been razed
to the ground by Llywelyn ap Jorwerth.[919] The bailey is large and
bean-shaped.

LAWHADEN, or Llanyhadein, or Lauwadein.--First mention in 1192.[920]
It afterwards became a palace of the bishops of St David's. There is no
motte, though the circular outline of the platform on which the fine
ruins of the castle stand, very much suggests a lowered motte.

HAVERFORDWEST.--First mentioned in the _Pipe Roll_ of 1214-1215, when
it was in the custody of the Earl of Pembroke. Although this castle is
now a gaol, and the whole site masked with gaol buildings, the motte
can still be seen distinctly from one side, though the keep which
stands upon it is blocked by buildings. The ditch which went round the
motte can also be traced. [H. W.]

NARBERTH.--This castle is first mentioned in 1115, when it was burnt
by the Welsh. Said to have been the castle of Stephen Perrot.[921] The
present ruins are entirely of the 13th century, and there is no motte;
but Lewis states that the first castle was in another site, between the
present town and Templeton; about which we have no information.

TENBY.--First mention in 1152. An important coast station. The small
and curious round keep is placed on the highest point of a small
island; it is a miniature copy of the keep of Pembroke, and was
probably built by one of the earls Marshall, not earlier than the 13th
century. There is no motte, nor was one needed in such a situation.


CASTLES OF CARDIGAN.

CARDIGAN Castle, or Aberteifi, has been so much transformed by the
incorporation of the keep into a modern house that nothing decisive can
be said about its original plan, but there is nothing to foreclose
the idea of a previous motte, and Speed's plan of 1611 seems to show
that the keep and the small ward attached to it were on a higher
elevation than the bailey. That the first castle was a wooden one is
rendered almost certain by the fact that Rhys ap Griffith, after having
demolished the previous castle, rebuilt it _with stone and mortar_,
in the reign of Henry II.[922] The Welsh chronicler speaks of this
castle as the key of all Wales, an exaggeration certainly, but it was
undoubtedly the most important stronghold of South Ceredigion. [H. W.]

CILGERRAN, or Dingeraint (Fig. 41).--This castle was certainly built
by Earl Roger;[923] a castle of great importance, in a magnificent
situation. Like nearly all the castles in our Welsh list, it was
repeatedly taken by the Welsh and retaken from them. The present
masonry is of the 13th century, but the original motte-and-bailey plan
is quite discernible. [H. W.] It was a connecting link between the
castles of Pembrokeshire and those of Cardigan, and stands near a road
leading directly from Tenby and Narberth to Cardigan.

ABERYSTWYTH, also Lampadarn Vaur, also Aberrheiddiol.[924] In 1109
Henry I. deposed Cadwgan, a Welsh prince who had purchased from the
king the government of Cardigan, and gave that country to Gilbert, son
of Richard, Earl of Clare, who took possession, and built a castle
"opposite to Llanbadarn, near the mouth of the river Ystwyth."[925]
This was undoubtedly the precursor of the modern castle of Aberystwyth,
but it is doubtful whether it was on the same site; the present ruins
are not opposite Llanbadarn. The castle was as important for the
defence of N. Cardigan as Cardigan Castle for the south. It was taken
at least seven times by the Welsh, and burnt at least five times. The
present ruins are not earlier than the time of Edward I., and there is
no motte or keep. [H. W.]

*BLAENPORTH, or Castell Gwythan (Fig. 41).--Also built by Gilbert de
Clare, and evidently placed to defend the main road from Cardigan to
Aberystwyth. The motte and bailey are still remarkably perfect, as
shown by the 25-inch Ordnance Map.

YSTRAD PEITHYLL.--Another of Gilbert de Clare's castles, as it was
inhabited by his steward. It was burnt by the Welsh in 1115,[926] and
is never mentioned again, but its motte and ditch still survive, with
some signs of a bailey, close to the little stream of the Peithyll,
near Aberystwyth. [H. W.]

CHASTELL GWALTER, or Llanfihangel, in Pengwern (Fig. 41).--Castle of
Walter de Bec, probably one of the barons of Gilbert de Clare. First
mentioned in 1137, when it was burned by the Welsh.[927] There is a
small but well-made motte and part of an adjoining bailey standing in a
most commanding position on a high plateau. The ditch of the motte is
excavated in the rock. [D. H. M.]

[Illustration: FIG. 41.--MOTTE-CASTLES OF SOUTH WALES. Cilgerran.
Blaenporth. Chastell Gwalter.]

*DINERTH.--Also burnt in 1137; restored by Roger, Earl of Clare,
in 1159, after which it underwent many vicissitudes.[928] Probably
originally a castle of the Clares. "In the grounds of Mynachty, in
the parish of Llanbadarn Tref Eglwys, is a small hill called Hero
Castell, probably the site of the keep of Dinerth Castle."[929] The
O.M. shows a small motte and bailey placed between two streams.

*CAERWEDROS, or Castell Llwyndafydd, also burned by the Welsh in
1137,[930] after which it is not mentioned again. "A very large moated
tumulus, with foundations of walls on the top."[931] Probably a Clare
castle.

*HUMPHREY'S CASTLE, now Castle Howel, from one of its Welsh conquerors.
The original name shows that it was built by a Norman, and it was
restored by Roger, Earl of Clare, in 1159.[932] A moated tumulus near
the river Clettwr marks the site of Humphrey's Castle.[933]

YSTRAD MEURUG, or Meyric, at the head of the valley of the Teifi,
and commanding the pass leading over into Radnorshire.--Built by
Gilbert de Clare when he reconquered Cardigan, and one of his most
important castles.[934] Its importance is shown by the fact that it
had a small stone keep, the date of which cannot now be determined,
as only the foundations remain, buried under sods. There is no motte,
and the bailey can only be guessed at by a portion of the ditch which
still remains on the N. side, and by two platforms which appear to be
artificially levelled. The castle is about three miles from the Sarn
Helen or Roman road through Cardigan.

*PONT Y STUFFAN, or Stephen's Bridge, near Lampeter.--Burnt by the
Welsh in 1138, and not again mentioned.[935] In the outskirts of
the town of Lampeter is--or was--a lofty moated tumulus (not shown
on O.M.), and traces of a quadrangular court.[936] As it is also
called Castell Ystuffan, it was probably built by Stephen, the Norman
constable of Cardigan. There appears to be another castle mound at
Lampeter itself, near the church. Lampeter was an important post on the
Roman road up the valley of the Teifi.

*NANT YR ARIAN.--This castle is only mentioned once, in the partition
of Cardigan and Pembroke which took place in 1216, during the most
disastrous part of John's reign.[937] There are two "castellau" marked
at Nant yr Arian in the N. of Cardiganshire in the O.M.; neither of
them look like mottes. This castle, as well as that of Ystrad Peithyll,
seems to have been placed to defend the road from Aberystwyth to
Llanidloes, which would be the chief highway between Shropshire and
Ceredigion.


CASTLES OF CARMARTHENSHIRE.

RHYD Y GORS, or Rhyd Cors.--We have no hesitation in adopting the
opinion of the late Mr Floyd, that this is another name for the castle
of Carmarthen.[938] As it and Pembroke were the only castles which
held out during the great Welsh revolt of 1096,[939] it is evident
that they were the two strongest and best defended places, therefore
the most important. Carmarthen also was a Roman city, and its walls
were still standing in Giraldus' time;[940] it was therefore the place
where one would expect to find a Norman castle. Now Carmarthen, along
with Cardiff and Pembroke, continued up till the final conquest of all
Wales to be the most important seat of English power in South Wales.
Moreover, Rhyd y Gors was a royal castle; we are expressly told that
it was built by William Fitz Baldwin, by the command of the king of
England.[941] Carmarthen also was a royal castle, and the only one in
South Wales at that date which belonged directly to the king. It was
temporarily abandoned after William Fitz Baldwin's death in 1096, and
afterwards Henry I. gave it into the custody of a Welshman, who also
had charge of Strath Towy; a passage which proves that Rhyd y Gors was
in that district. It was restored by Richard Fitz Baldwin in 1104,[942]
and is mentioned for the last time in 1105. After that the castle of
Carmarthen, which has not been mentioned before, begins to appear, and
its importance is clear from the continual references to it. Placed
as it is on a navigable river, at the entrance of the narrower part
of the vale of Towy, and on the Roman road from Brecon to St David's,
its natural position must have marked it as a fit site for a royal
castle. The castle is now converted into a gaol, and disfigured in the
usual way; yet the ancient motte of William Fitz Baldwin still remains,
partly inside and partly outside the walls. It is crowned with a stone
revetment which Colonel Morgan believes to have been erected at the
time of the Civil War, to form a platform for guns.[943] The bailey is
rectangular and covers about 2 acres. The motte is placed at one corner
of it, on the line of the walls. On the outside it is now built over
with poor cottages; but the site of the ditch can still be traced.

*LLANDOVERY, or Llanymdyfri, or the castle of Cantrebohhan.--It is
referred to in the _Pipe Rolls_ of 1159-1160 by the latter name, which
is only a Norman way of spelling Cantref Bychan, the little cantref
or hundred, of which this castle was the head.[944] It was then in
royal custody, and Henry II. spent nearly £60 on its works. But it had
originally belonged to Richard Fitz Pons, one of the barons of Bernard
de Neufmarché, and the fact that he held the key of this cantref goes
to prove that it was from Brecknock that the Normans advanced into
northern Carmarthenshire. The castle is first mentioned in the _Brut_
in 1115, when Griffith ap Rhys burnt the bailey, but could not take
the keep on the motte.[945] It does not appear to have been long in
English hands after 1159, but its alternations were many. The 25-inch
O.M. shows an oval motte, carrying some fragments of masonry, to which
is attached a roughly quadrangular bailey. This was one of the many
castles by which the Normans held Strath Towy.

LLANSTEPHAN.[946]--This castle stands in a splendid situation at
the mouth of the Towy, and was doubtless built to secure a maritime
base for Carmarthen. The motte is of unusual size, semicircular in
shape, one side being on the edge of the cliff; it measures 300 feet
by 200 in the centre of the arc.[947] Such a size allowed all the
important parts of the castle to be built on the motte; but there was
a rectangular bailey attached, which is only imperfectly shown on the
O.M.; the scarp is in reality well marked on all sides, and the ditch
separating it from the motte is a very deep one. [H. W.] The towers
that now crown the motte are not earlier than the year 1256, when the
castle was destroyed by Llywelyn.[948]

DINEVOR, or Dinweiler.--Most Welsh writers associate Dinevor with the
ancient residence of the kings of South Wales, but there appears to
be some doubt about this, as the place is not mentioned before the
12th century.[949] Anyhow the castle was certainly the work of Earl
Gilbert, as the _Brut_ itself tells us so.[950] In 1162 it was taken
by Rhys ap Griffith, the able prince who attempted the consolidation
of South Wales, and who was made Justiciar of that province by Henry
II. It continued in Welsh hands, sometimes hostile, sometimes allied,
till it was finally taken by the English in 1277. The existing ruins
are entirely of the 13th century, but the plan certainly suggests a
previous motte and bailey, the motte having probably been lowered to
form the present smaller ward, whose walls and towers appear to be of
Edward I.'s reign. The small bailey attached to this ward is separated
from it by a ditch cut through the headland on which the castle stands.

KIDWELLY (Cydweli).--This castle, though in Carmarthen, was not founded
by the conquerors from Brecknock, but by Normans from Glamorgan or
Gower. Kidwelly was first built by William de Londres, in 1094.[951]
The present castle shows no trace of this early origin, but is a fine
specimen of the keepless pattern introduced into England in the 13th
century.[952] There is no motte.

LAUGHARNE, or Talycharne.--Also called Abercorran, being at the point
where the little river Corran flows into the estuary of the Taff. In
1113 this castle belonged to a Norman named Robert Courtmain.[953] The
ancient features of the plan have been obliterated by transformation
first into an Edwardian castle, then into a modern house. There is of
course no motte. [H. W.]

*YSTRAD CYNGEN.--This must, we think, be the same as ST CLEARS, which
stands in the Cynen valley, near its junction with the Taff. Welsh
writers identify St Clears with the castle of Mabudrud, the name of the
_commot_ in which it stands. First mentioned in 1154.[954] There is no
notice of its origin, but the fact that a Cluniac priory existed in the
village, which was a cell of St Martin des Champs at Paris, points to
a Norman founder, and renders an 11th century date probable. It was a
motte-and-bailey castle, of which the earthworks remain.[955]

*NEWCASTLE EMLYN.--This castle does not appear to have received the
name of "the new castle of Emlyn" till after Edward I.'s conquest.[956]
The new castle, which is quite Edwardian, was probably built on a
different site to the old, as "on the other side of the bridge is a
considerable mount, of a military character, which must have commanded
the river. It may have been the original strong post occupied by the
Normans."[957] In the 12th century _Pipe Rolls_ compensation is paid
to William FitzGerald for many years "as long as Rhys ap Griffith
holds the castle of Emlyn," which points to Gerald, the Seneschal of
Pembroke, or his family, as its founders. It is on the very border
of Carmarthenshire and Cardiganshire, defending the main road from
Carmarthen to Cardigan.

LLANEGWAD.--This castle is only once mentioned, in the _Brut_, under
the year 1203, when it was taken by the Welsh. A small motte, called
locally Pen y Knap, with an earthen breastwork round the top, is still
standing about a mile from the church of Llanegwad, and is all that is
left of this castle. The position commands a fine view over the Towy
valley, and it is noteworthy that it stands very near the supposed
Roman road from Brecon to Carmarthen. [H. W.]

*LLANGADOG.--This castle also does not appear till 1203; it was razed
or burnt at least thrice in five years.[958] A mound of earth on the
banks of the Sawddwy River, near where the Roman road from Brecon is
supposed to have reached the Towy valley, is all that remains of
it.[959] Lewis says that it stands in a large oval entrenchment, and
that the motte is of natural rock, scarped conically, and deeply moated.


CASTLES IN BRECKNOCKSHIRE.

BRECON, or Aberhonddu, the seat of Bernard de Neufmarché himself.--A
charter of Bernard's mentions the castle.[960] It seems to have been a
particularly strong place, as we do not hear of its having been burnt
more than once. The newer castle of Brecon is evidently of the time
of Edward I., but across the road the old motte of Bernard is still
standing, and carries the ruins of a shell wall, with a gatehouse
tower.[961] A portion of the bank and ditch of the bailey remains;
the whole is now in a private garden. The situation is a strong one,
between the Usk and the Honddu. Brecon of course was a burgus, and part
of the bank which fortified it remains.

[Illustration: FIG. 42.--MOTTE-CASTLES OF SOUTH WALES. Builth. Gemaron.
Payn's Castle.]

BUILTH, on the upper Wye, _alias_ Buallt (Fig. 42).--A remarkably
fine motte and bailey, presenting some peculiarities of plan. It
is not mentioned till 1210,[962] but it has been conjectured with
great probability that it was one of the castles built by Bernard de
Neufmarché when he conquered Brecknock.[963] It was refortified by
John Mortimer in 1242,[964] probably in stone, as in the account of its
destruction by Llywelyn in 1260 it is said that "not one stone was left
on another."[965] Nevertheless when Edward I. rebuilt it the towers on
the outer wall appear to have been of wood.[966] Mr Clark states that
there are traces of masonry foundations and small portions of a wing
wall. The bailey of this castle consists of a rather narrow platform,
divided into two unequal portions by a cross ditch which connects the
ditch of the motte with that of the bailey. The ditch round the motte
is of unusual breadth, being 120 feet broad in the widest part. The
whole work is encircled by an outer ditch of varying breadth, being 100
feet wide on the weakest side of the work, and by a counterscarp bank
which appears to be still perfect. The entrance is defended by four
small mounds which probably cover the remains of towers.[967] The area
of the two baileys together is only 1 acre. [D. H. M.]

*HAY, or Tregelli.--The earliest mention of this castle is in a charter
of Henry I.[968] The present castle of Hay is of late date, but Leland
tells us that "not far from the Paroche Chirch is a great round Hille
of Yerth cast up by Men's Hondes."[969] It is shown on the 25-inch
O.M., and so is the line of the borough walls.

*TALGARTH.--Mentioned in a charter of Roger, Earl of Hereford, not
later than 1156.[970] A 13th-century tower on a small motte is still
standing, and can be seen from the railway between Brecon and Hereford.


CASTLES OF RADNORSHIRE.

*RADNOR, or Maes Hyvaidd.--Though this castle is not mentioned in
the _Brut_ till 1196, when it was burnt by Rhys ap Griffith, it must
have been built by the Normans at a very early period. The English
had penetrated into the Radnor district even before the Norman
Conquest,[971] and the Normans were not slow to follow them. A charter
of Philip de Braose is granted at "Raddenoam" not later than 1096.[972]
There are mottes both at Old and New Radnor, towns three miles distant
from each other, so that it is impossible to say which was the Maes
Hyvaidd of the _Brut_. Both may have been originally De Braose castles,
but New Radnor evidently became the more important place, and has
massive remains in masonry. The town was a _burgus_.

*GEMARON, or Cwm Aron (Fig. 42).--Near Llandewi-Ystrad-denny. The
_Brut_ mentions its repair by Hugh Mortimer in 1145.[973] The 6-inch
O.M. shows a square central bailey of 1 acre, containing some remains
of masonry, lying between an oblong motte in the S. and an outer
enclosure on the N., the whole being further defended by a high
counterscarp bank on the W. It commands a ford over the river Aran.
There is no village attached to it.

*MAUD'S CASTLE, otherwise Colwyn or Clun.[974]--A ditched motte with
square bailey on the left bank of the river Edwy, near the village of
Forest Colwyn. The statement that this castle was _repaired_ in 1145
shows that it must have been older than the time of Maude de Braose,
from whom it is generally supposed to have taken its name. It was
rebuilt by Henry III. in 1231.[975]

*PAYN'S CASTLE, otherwise "the castle of Elvael."--First mentioned
in 1196, when it was taken by Rhys ap Griffith. This is also a
motte-castle (and an exceptionally fine one), placed on a road leading
from Kington in Hereford to Builth. Rebuilt _in stone_ by Henry III. in
1231.[976] (Fig. 42.)

*KNIGHTON, in Welsh Trefclawdd.--First mentioned in the _Pipe Roll_ of
1181. The motte still remains, near the church. There is another motte
just outside the village, called Bryn y Castell. It may be a siege
castle.

*NORTON.--First mentioned in the _Pipe Roll_ of 1191. A motte remains
close to the church, and two sides of a bailey which ran down to the
Norton brook.

*BLEDDFA, the Bledewach of the _Pipe Roll_ of 1195-1196, when £5 was
given to Hugh de Saye _ad firmandum castellum_, an expression which
may mean either building or repairing. An oval motte, and traces of a
bailey, are marked in the 6-inch O.M.

TYNBOETH, _alias_ Dyneneboth, Tinbech,[977] and Llananno.--First
mentioned in _Pipe Roll_ of 1196-1197. There is a fine large motte
in a commanding situation, and a crescent-shaped bailey, now marked
only by a scarp. There are some remains of masonry, and the castle was
evidently an important one. It is first mentioned in the _Pipe Roll_
of 1196, and it occurs in lists of the Mortimer castles in the 14th
century.[978] It is not far from two fords of the river Ithon. [H. W.]

These four castles are not mentioned in the _Brut y Tywysogion_, though
the _Annales Cambriæ_ mentions the capture of Bleddfa, Knighton, and
Norton by the Welsh in 1262. They all command important roads. Knighton
and Norton were boroughs.

[Illustration: FIG. 43.--MOTTE-CASTLES OF SOUTH WALES. Cardiff.
Loughor.]


CASTLES OF GLAMORGANSHIRE.

_Cardiff_ (Fig. 43).--The first castle of Cardiff was certainly a
wooden one; its lofty mound still remains. It is placed inside a
Roman station, and the south and west walls of the castle bailey
rest on Roman foundations, "but do not entirely coincide with those
foundations."[979] The Roman fort was probably ruinous when Robert Fitz
Hamon placed his first castle there, as on the N. and E. sides the
bailey is defended by an earthbank, in which the remains of a Roman
wall have been found buried. The area of the Roman Castrum was about
8-1/4 acres, and evidently the Normans found this too large, as they
divided it by a cross wall, which reduces the inner fort to about 2
acres. The motte has its own ditch. The position of Cardiff was a very
important base, not only as a port near Bristol, but as a point on
the probably Roman road which connected Gloucester with Carmarthen and
beyond.[980]

The lands of Robert Fitz Hamon, in the next generation, passed into the
hands of Robert, the great Earl of Gloucester, Henry I.'s illegitimate
son. He was a great castle-builder, and it is probable that the first
masonry of Cardiff Castle was his work.[981]

NEWCASTLE BRIDGEND.--This castle and the three which follow are all
situated on or near the "Roman" road from Cardiff to St David's, of
which we have already spoken. There were two castles at Bridgend, the
Old Castle and the New Castle, from which the town takes its name. The
site of the former is now too much cut up for any definite conclusions
about it; the site of the latter has been converted into market
gardens, but a motte is still standing in one corner with the ruins
of a tower upon it. [H. W.] This castle is not noticed either by the
_Brut_ or the Aberpergwm version; the earliest mention known to us is
in the _Pipe Roll_ of 1184, at a time when the castles of the Earl of
Gloucester were in royal custody, and this appears to have been one of
them.

KENFIG.--This castle is close to the "Roman" road. The Aberpergwm
_Brut_ says that it was one of the castles of Robert Fitz Hamon, and
states that in 1092 it was rebuilt "stronger than ever before, for
castles prior to that were built of wood." This is a good specimen
of the mixture of truth and error to be found in this 16th century
MS. There is little doubt that all the first castles of the Normans
in Wales were built of wood; but it is extremely unlikely that any
wooden keep was replaced by a stone one as early as 1092. The town and
castle of Kenfig are now almost entirely buried in sand-drifts, but
the top of the motte, with some fragments of masonry upon it, is still
visible. [H. W.][982] The note in the _Pipe Rolls_ of the repair of the
_palicium_ of this castle shows that the bailey wall at any rate was
still of wood in 1183. Even as late as 1232 the keep was only defended
by a ditch and hedge; yet it withstood an assault from Llywelyn ap
Jorwerth.[983] The bailey is said to contain 11 acres, a most unusual
size. Kenfig was a borough in Norman times, and it is possible that
this large bailey was the original borough, afterwards enlarged in
mediæval times. There is evidence that there were burgage tenements
within the bailey.[984]

ABERAVON.--The Aberpergwm MS. says that Fitz Hamon gave Aberavon to the
son of the Welsh traitor who had called him into Glamorgan. At a later
period, however, we find it in Norman hands. The site of the castle
has been entirely cleared away, but it had a motte, which is still
remembered by the older inhabitants. [H. W.][985] It is not mentioned
in the _Brut_ before 1152, when it was attacked and burnt by Rhys ap
Griffith.

*NEATH.--The site of the first castle of Neath was given by Richard de
Granville, its owner, to the abbey of Neath, which he had founded.[986]
About the year 1111, according to the Aberpergwm _Brut_, Richard
returned from the Holy Land, bringing with him a Syrian architect, well
skilled in the building of monasteries, churches, and castles, and by
him we may presume, a new castle was built on the other side of the
river, though the present castle on that site is clearly of much later
date. The monks of course destroyed all vestiges of the first (probably
wooden) castle.

*REMMI, or Remni.--Of this castle there is only one solitary mention,
in the _Pipe Roll_ of 1184. The name seems to indicate the river
Rhymney, which is the boundary between Glamorgan and Monmouth. We
are unable to find any castle site so near the Rhymney as Ruperra,
where Clark mentions a fine motte.[987] But we do not venture on this
identification without further information.[988]


CASTLES OF GOWER.

*SWANSEA, or Abertawy.--This was the castle of Henry Beaumont, the
conqueror of Gower. The present castle is comparatively modern. It is
inside the town; but there used to be a moated mound outside the town,
which was only removed in 1804. It seems probable to us that this was
the original castle of Beaumont.[989] That this first castle had a
motte is suggested by the narrative in the _Brut_ which tells how
Griffith ap Rhys burnt the outworks in 1115, but was unable to get at
the tower.[990]

*LOUGHOR, or Aberllychor (Fig. 43).--Also built by Henry Beaumont.
The mound of the castle still remains, with a small square keep on
top. There was formerly a shell wall also. The place of a bailey
was supplied by a terrace 15 feet wide.[991] The four castles last
mentioned are all at the mouths of rivers, as well as on an ancient (if
not Roman) coast road.

*LLANDEILO TALYBONT, or Castell Hu.--Only mentioned once in the _Brut_,
under 1215, as the castle of Hugh de Miles. A moated mound with a
square bailey and no masonry still remains.[992] It commands the river
Loughor, which is still navigable up to that point at high tides.[993]
On the opposite side of the river is another motte and bailey, called
Ystum Enlle. Possibly there was a ford or ferry at this point, which
these castles were placed to defend.[994]

OYSTERMOUTH, a corruption of Ystum Llwynarth.--First mentioned in the
older _Brut_ in 1215, when it was burnt by Rhys Grug. The later version
says it was built by Beaumont in 1099. The castle stands on a natural
height, fortified artificially by a motte, which is of great size.
There is a small bailey below to the N.E., and a curious small oval
embankment thrown out in the rear of the castle towards the N.W. The
architecture of this magnificent castle is all of the Edwardian style,
and as the castle was burnt down by Rhys ap Meredith in 1287, it is
probable that only wooden structures stood on this site until after
that date. The castle is in a fine situation overlooking the Bay of
Swansea. [H. W.]

       *       *       *       *       *

We have now completed our list of the Norman castles built in Wales
which are known to history. It must not be supposed, however, that we
imagine this to be a complete list of all the Norman castles which
were ever erected in Wales. The fact that several in our catalogue are
only once mentioned in the records makes it probable that there were
many others which have never been mentioned at all. In this way we may
account for the many mottes which remain in Wales about which history
is entirely silent. As there was scarcely a corner in Wales into
which the Normans did not penetrate at some time or other, it is not
surprising if we find them in districts which are generally reckoned
to be entirely Welsh. But there is another way of accounting for them;
some of them may have been built by the Welsh themselves, in imitation
of the Normans. As the feudal system and feudal ideas penetrated more
and more into Wales, and the Welsh princes themselves became feudal
homagers of the kings of England, it was natural that the feudal castle
should also become a Welsh institution, especially as it was soon
found to be a great addition to the chieftain's personal strength. The
following castles are stated in the _Brut_ to have been built by the
Welsh.[995]

1113. *CYMMER, in Merioneth.--Built by Uchtred ap Edwin, whose name,
as we have already remarked, suggests an English descent. Near Cymmer
Abbey the motte or _tomen_ remains.

*CYNFAEL, in Merioneth, near Towyn.--Built by Cadwalader, son of
Griffith ap Cynan, on whose behalf Henry II. undertook his first
expedition into Wales, and who was at that time a protégé of the
Anglo-Normans. Clark gives a plan of this motte-castle in _Arch.
Camb._, 4th ser., vi., 66.

1148. *YALE, in Denbigh = Llanarmon.--Said to have been built by Owen
Gwynedd, but here, as we have said, an earlier Norman foundation seems
probable (see p. 272).

1148. LLANRHYSTYD, in Cardigan.--Also built by Cadwalader, who was then
establishing himself in Cardigan. Probably the motte and bailey called
Penrhos, or Castell Rhos, to the east of Llanrhystyd village. [H. W.]

1155. ABERDOVEY.--Built by Rhys ap Griffith to defend Cardigan against
Owen, Prince of Gwynedd. It must therefore have been on the Cardigan
shore of the Dovey, and not at the present town of Aberdovey, which
is on the Merioneth shore. And in fact, on the Cardigan shore of the
estuary, about two miles west of Glandovey Castle, there is a tumulus
called Domenlas (the green tump), which was very likely the site of
this castle of Rhys.[996]

1155. CAEREINION.--Built by Madoc of Powys, who was then a homager of
Henry II. Remains of a motte near the church; the churchyard itself
appears to be the former bailey. About a mile off is a British camp
called Pen y Voel, which _may_ have been the seat of the son of
Cunedda, who is said to have settled here. [H. W.]

*WALWERN, or Tafolwern, near Llanbrynmair, in Montgomery, may have been
a Welsh castle. It is first mentioned in 1163, when Howel ap Jeuav
took it from Owen Gwynedd, who may have been its builder. The motte is
marked in the O.M. on a narrow peninsula at the junction of two streams.

1169. *ABEREINON, in Cardigan.--Built by Rhys ap Griffith, Henry II.'s
Justiciar of South Wales. "A circular moated tumulus, now called Cil y
Craig."[997] (It is marked on the 25-inch O.M.)

1177. *RHAIDR GWY.--Also built by Rhys ap Griffith, no doubt as a
menace to Powys, as this castle was afterwards sorely contested. It is
a motte-and-bailey castle, the motte being known as Tower Mount.[998]

       *       *       *       *       *

All these castles are of the motte-and-bailey type, and prove the
adoption by the Welsh of Norman customs.[999] It will be noticed that
in the first instances they were built by men who were specially under
Norman influences. But probably the fashion was soon more widely
followed, although these are the only recorded cases.

The contribution made by the castles of Wales to the general theory
of the origin of mottes in these islands is very important. Leaving
out the seven castles attributed to the Welsh, we find that out of
seventy-one castles built by the Normans, fifty-three, or very nearly
three-fourths, still have mottes; while in the remaining eighteen,
either the sites have been so altered as to destroy the original plan,
or there is a probability that a motte has formerly existed.




CHAPTER X

MOTTE-CASTLES IN SCOTLAND


The Scottish historians of the 19th century have amply recognised the
Anglo-Norman occupation of Scotland, which took place in the 11th and
12th centuries, ever since its extent and importance were demonstrated
by Chalmers in his _Caledonia_. Occupation is not too strong a word to
use, although it was an occupation about which history is strangely
silent, and which seems to have provoked little resistance except in
the Keltic parts of the country. But it meant the transformation of
Scotland from a tribal Keltic kingdom into an organised feudal state,
and in the accomplishment of this transformation the greater part of
the best lands in Scotland passed into the hands of English refugees or
Norman and Flemish adventurers.

The movement began in the days of Malcolm Canmore, when his English
queen, the sainted Margaret, undoubtedly favoured the reception
of English refugees of noble birth, some of whom were her own
relations.[1000] Very soon, the English refugees were followed by
Norman refugees, who had either fallen under the displeasure of the
king of England, like the Montgomeries, or were the cadets of some
Norman family, wishful to carve out fresh fortunes for themselves, like
the Fitz Alans, the ancestors of the Stuarts. The immigration continued
during the reign of the sons of Margaret, but seems to have reached its
culminating point under David I. (1124-1153).

David, as Burton remarks, had lived for sixteen years as an affluent
Anglo-Norman noble, before his accession to the Scottish crown, being
Earl of Huntingdon in right of his wife, the daughter of Simon de
Senlis, and granddaughter, through her mother, of Earl Waltheof.
David's tastes and sympathies were Norman, but it was not taste
alone which impelled him to build up in Scotland a monarchy of the
Anglo-Norman feudal type. He had a distinct policy to accomplish; he
wished to do for Scotland what Edward I. sought to do for the whole
island, to unite its various nationalities under one government, and
he saw that men of the Anglo-Norman type would be the best instruments
of this policy.[1001] It mattered little to him from what nation he
chose his followers, if they were men who accepted his ideas. Norman,
English, Flemish, or Norse adventurers were all received at his
court, and endowed with lands in Scotland, if they were men suitable
for working the system which he knew to be the only one available
for the accomplishment of his policy. And that system was the feudal
system. He saw that feudalism meant a higher state of civilisation
than the tribalism of Keltic Scotland, and that only by the complete
organisation of feudalism could he carry out the unification of
Scotland, and the subjugation of the wild Keltic tribes of the north
and west.[1002]

The policy was successful, though it was not completely carried out
until Alexander III. purchased the kingdom of the Isles from the King
of Norway in 1266. The sons of David, Malcolm IV., and William the Lion
were strong men who doughtily continued the subjugation of the Keltic
parts of Scotland, and distributed the lands of the conquered among
their Norman or Normanised followers. The struggle was a severe one;
again and again did the North rebel against the yoke of the House of
Malcolm. In Moray the Keltic inhabitants were actually driven out by
Malcolm IV., and the country colonised by Normans or Flemings.[1003]
The same Malcolm led no less than three expeditions against Galloway,
where in spite of extensive Norse settlements on the coast, the mass of
the inhabitants appear to have been Keltic.[1004]

We know very little about the details of this remarkable revolution,
because Scotland had no voice in the 12th century, none of her
chroniclers being earlier than the end of the 14th century. As regards
the subject which concerns this book, the building of castles, there
are only one or two passages which lift the veil. A contemporary
English chronicler, Ailred of Rievaulx, in his panegyric of David I.,
says that David decorated Scotland with castles and cities.[1005] In
like manner Benedict of Peterborough tells us that when William the
Lion was captured by Henry II.'s forces in 1174, the men of Galloway
took the opportunity to destroy all the castles which the king had
built in their country, expelling his seneschals and guards, and
killing all the English and French whom they could catch.[1006] Fordun
casually mentions the building of two castles in Ross by William the
Lion; and once he gives us an anecdote which is a chance revelation
of what must have been going on everywhere. A certain English knight,
Robert, son of Godwin, whose Norman name shows that he was one of the
Normanised English, tarried with the king's leave on an estate which
King Edgar had given him in Lothian, _and while he was seeking to build
a castle there_, he was attacked by the men of Bishop Ranulf of Durham,
who objected to a castle being built so near the English frontier.[1007]

But even if historians had been entirely silent about the building of
castles in Scotland, we should have been certain that it must have
happened, as an inevitable part of the Norman settlement. Robertson
remarks that the Scots in the time of David I. were still a pastoral
and in some respects a migratory people, their magnates not residing
like great feudal nobles in their own castles, but moving about
from place to place, and quartering themselves upon the dependent
population. There is in fact no reason for supposing that the Keltic
chiefs of Scotland built castles, any more than those of Wales or
Ireland.[1008] But the feudal system must very soon have covered
Scotland with castles.

The absence of any stone castles of Norman type has puzzled Scottish
historians, whose ideas of castles were associated with buildings in
stone.[1009] In 1898 Dr Christison published his valuable researches
into the _Early Fortifications of Scotland_, in which for the first
time an estimate was attempted of the distribution of Scottish
_motes_,[1010] and their Norman origin almost, if not quite, suspected.
His book was quickly followed by Mr George Neilson's noteworthy paper
on the "Motes in Norman Scotland,"[1011] in which he showed that the
wooden castle is the key which unlocks the historians' puzzle, and that
the motes of Scotland are nothing but the evidence of the Norman feudal
settlement.

Two important points urged in Mr Neilson's paper are the feudal and
legal connection of these motes. He has given a list of mottes which
are known to have been the site of the "chief messuages" of baronies
in the 13th and 14th centuries, and has collected the names of a great
number which were seats of justice, or places where "saisine" of a
barony was taken, not because they were moot-hills, but because the
administration of justice remained fixed in the ancient site of the
baron's castle. "The doctrine of the chief messuage, which became of
large importance in peerage law, made it at times of moment to have on
distinct record the nomination of what the chief messuage was, often
for the imperative function of taking _sasine_. In many instances the
_caput baroniæ_, or the court or place for the ceremonial entry to
possession, is the 'moit,' the 'mothill,' the 'auld castell,' the 'auld
wark,' the 'castellsteid,' the 'auld castellsteid,' the 'courthill,'
or in Latin _mons placiti_, _mons viridis_, or _mons castri_."[1012]
In certain places where two mottes are to be found, he was able to
prove that two baronies had once had their seats. Another point which
Mr Neilson worked out is the relation of bordlands to mottes. Bordland
or borland, though an English word, is not pre-Conquest; it refers
to "that species of demesne which the lord reserves for the supply
of his own table." It is constantly found in the near proximity of
mottes.[1013]

The following is a list of thirty-eight Anglo-Norman or Normanised
adventurers settled in Scotland, on whose lands mottes are to be found.
The list must be regarded as a tentative one, for had all the names
given by Chalmers been included, it would have been more than doubled.
But the difficulties of obtaining topographical information were so
great that it has been judged expedient to give only the names of those
families who are known to have held lands, and in most cases to have
had their principal residences, in places where mottes are or formerly
were existing.[1014]

ANSTRUTHER.--William de Candela obtained the lands of Anstruther, in
Fife, from David I. His descendants took the surname of Anstruther. The
"Mothlaw" of Anstruther is mentioned in 1590.[1015] "At the W. end of
the town there is a large mound, called the Chester Hill, in the middle
of which is a fine well." (N. S. A., 1845.) The well is an absolute
proof that this was the site of a castle.

AVENEL.--Walter de Avenel held Abercorn Castle and estate, in
Linlithgow, in the middle of the 12th century. The castle stood on a
green mound (N. S. A.) which is clearly marked in the O.M.

BALLIOL.--The De Bailleul family had their seat at Barnard Castle, in
Durham, after the Conquest. They obtained lands in Galloway from David
I., and had strongholds at Buittle, and Kenmure, in Kirkcudbright. At
Buittle the site of the castle exists, a roughly triangular bailey with
a motte at one corner;[1016] and at Kenmure the O.M. clearly shows a
motte, as does the picture in Grose's _Antiquities of Scotland_. The
terraces probably date from the time when the modern house on top was
built.

BARCLAY.--The De Berkeleys sprang from the De Berkeleys of England,
and settled in Scotland in the 12th century. Walter de Berkeley was
Chamberlain of Scotland in 1165; William the Lion gave him the manor
of Inverkeilor, in Forfarshire; there he built a castle, on Lunan Bay.
"An artificial mound on the west side of the bay, called the Corbie's
Knowe, bears evident marks of having been a castle long previous to
the erection of Redcastle." (N. S. A.) The family also had lands in
what is now Aberdeenshire, and at Towie, in the parish of Auchterless,
they had a castle. "Close to the church of Auchterless there is a small
artificial eminence of an oval shape, surrounded by a ditch, which is
now in many places filled up. It still retains the name of the Moat
Head, and was formerly the seat of the baronial court." (N. S. A.; N.;
C.)

BRUCE.--The De Brus held lands in North Yorkshire at the time of
the Domesday Survey. David I. gave them the barony of Annan, in
Dumfriesshire. The original charter of this grant still exists in the
British Museum, witnessed by a galaxy of Norman names.[1017] Their
chief castles were at Annan and Lochmaben. At Annan, near the site of
a later castle, there is still a motte about 50 feet high, with a vast
ditch and some traces of a bailey (N.), called the Moat (N. S. A.). The
"terras de Moit et Bailyis, intra le Northgate," are mentioned in 1582.
South of the town of Lochmaben, on the N.W. side of the loch, is a fine
motte called Castle Hill, with some remains of masonry, which is still
pointed out as the original castle of the Bruces.[1018] (G.) The fine
motte and bailey at Moffat must also have been one of their castles, as
Moffat was one of their demesne lands. (Fig. 44.)

CATHCART.--Name territorial. Rainald de Cathcart witnesses a charter
(in the Paisley Register) in 1179. Near the old castle of Cathcart,
Lanark, is "an eminence called Court Knowe." (N. S. A.) As Mr Neilson
has shown, these court knowes and court hills are generally disused
mottes. The name Rainald is clearly Norman.

CHEYNE.--This family is first known in 1258, but had then been long
settled in Scotland, and were hereditary sheriffs of Banffshire.
Chalmers only mentions their manor of Inverugie, in Aberdeenshire.
Behind the ruins of Inverugie Castle rises a round flat-topped hill,
which was the Castle Hill or Mote Hill of former days. (N. S. A.)

COLVILLE.--Appears in Scotland in the reign of Malcolm IV., holding
the manors of Heton and Oxnam, in Roxburgh. About 1/4 mile from Oxnam
(which was a barony) is a moated mound called Galla Knowe. (O.M., C.,
and N.) Hailes identified the castle in Teviotdale, captured and burnt
by Balliol in 1333, with that of Oxnam.[1019] Le Mote de Oxnam is
mentioned in 1424 (N.).

[Illustration: FIG. 44.--SCOTTISH MOTTE-CASTLES. Annan. Moffat. Duffus.
Liddesdale.]

CUMYN, or COMYN.--The first of this family came to Scotland as the
chancellor of David I.[1020] First seated at Linton Roderick, in
Roxburghshire, where there is a rising ground, surrounded formerly
by a foss, the site of the original castle; (G.) a description
which seems to suggest a motte. William the Lion gave the Cumyns
Kirkintilloch in Dumbarton, and we afterwards find them at Dalswinton
in Dumfriesshire, and Troqueer in Kirkcudbright. At Kirkintilloch the
O.M. shows a square mount concentrically placed in a square enceinte.
The enclosure was apparently one of the forts on the wall of Agricola,
but the writer on Kirkintilloch in the N. S. A. suspected that it had
been transformed into a castle by the Cumyns. At Dalswinton the O.M.
shows a motte, and calls it the "site of Cumyn's Castle." At Troqueer,
"directly opposite the spot on the other side the river where Cumyn's
Castle formerly stood is a mote of circular form and considerable
height." (N. S. A.) The Cumyn who held Kirkintilloch in 1201, was made
Earl of Buchan, and held the vast district of Badenoch, or the great
valley of the Spey. The N. S. A. gives many descriptions of remains
in this region which are suggestive of motte-castles; we can only
name the most striking: Ruthven, "a castle reared by the Comyns on a
green conical mound on the S. bank of the Spey, thought to be partly
artificial," now occupied by ruined barracks; Dunmullie, in the parish
of Duthill, where "there can be traced vestiges of a motte surrounded
by a ditch, on which, according to tradition, stood the castle of the
early lords"; Crimond, where Cumyn had a castle, and where there is
a small round hill called Castle Hill; and Ellon, where the Earl of
Buchan had his head court, on a small hill which has now disappeared,
but which was anciently known as the moot-hill of Ellon. Saisin of the
earldom was given on this hill in 1476. (N. S. A.)

CUNNINGHAM.--Warnebald, who came from the north of England, was a
follower of the Norman, Hugh de Morville, who gave him the lands of
Cunningham, in Ayrshire, from which the family name was taken. In the
parish of Kilmaurs, which is in the district of Cunningham, there is a
"mote," which may have been the castle of Warnebald; at any rate the
original manor place of Cunningham was in this parish. It is of course
possible that this motte may have been originally a De Morville castle.

DOUGLAS.--Name territorial; progenitor was a Fleming, who received
lands on the Douglas water, in Lanark, in the middle of the 12th
century. In the park of Douglas, to the east of the modern castle, is
a mound called Boncastle, but we are unable to state certainly that it
is a motte. Lag Castle, in the parish of Dunscore, "has a moat or court
hill a little to the east." (N. S. A.: shown in Grose's picture.) It
must have been originally Douglas land, as in 1408 it was held by an
armour-bearer of Douglas.

DURAND.--Clearly a Norman name, corrupted into Durham. The family were
seated at Kirkpatrick Durham in the 13th century. There is or was a
motte at Kirkpatrick.[1021]

DURWARD.--This family was descended from Alan de Lundin, who was
dur-ward or door-keeper to the king about 1233. They possessed a wide
domain in Aberdeenshire, and had a castle at Lumphanan, where Edward I.
stayed in 1296. There is a round motte in the Peel Bog at Lumphanan,
surrounded by a moat, which was fed by a sluice from the neighbouring
burn. There were ruins in masonry on the top some hundred years ago.
The writer of the N. S. A. account of this place, with remarkable
shrewdness, conjectures that a wooden castle on this mound was the
ancient residence of the Durwards, superseded in the 15th century by
a building of stone, and that it has nothing to do with Macbeth, whose
burial-place is said to be a cairn in the neighbourhood.[1022]

FITZ ALAN.--This is the well-known ancestor of the House of Stuart,
Walter, a cadet of a great Norman family in Shropshire, who is said to
have obtained lands in Scotland in Malcolm Canmore's time. Renfrew was
one of his seats, and Inverwick, in Haddington, another. Renfrew Castle
is entirely destroyed, but the description of the site, on a small
hill, ditched round, called Castle Hill, strongly suggests a motte. The
keep of Inverwick stands on a natural motte of rock.[1023] Dunoon was
one of their castles, near to which "stood the Tom-a-mhoid, or Hill of
the court of justice" (G.), possibly an ancient motte.[1024] Dunoon
Castle, however, itself stands on a motte, partly artificial and partly
carved out of a headland. (N.)

FLEMING.--There were many Flemings among the followers of David I., and
eventually the name stuck to their descendants as a surname. Baldwin
the Fleming obtained lands at Biggar, in Lanarkshire. There is a motte
at the west end of the town of Biggar, 36 feet high. Biggar was the
head of a barony. (N. S. A. and N.) Colban the Fleming settled at
Colbantown, now Covington, Lanarkshire, where there is a motte (N.).
Robert the Fleming has left a well-preserved oblong motte at Roberton,
in Lanark, which was a barony, and where the _moit_ was spoken of in
1608. (N.)

GRAHAM.--Came from England under David I., and received lands in
Lothian. A Graham was lord of Tarbolton, in Ayrshire, in 1335, so it
is possible that the motte at that place, on which stood formerly the
chief messuage of the barony of Tarbolton, was one of their castles (N.
S. A.), but it may have been older.

HAMILTON.--It is not certain that the Hamiltons came to Scotland before
1272. King Robert I. gave them the barony of Cadzow, Lanark, which had
originally been a royal seat. In Hamilton Park there is a mote hill,
which was the site of the chief messuage of this barony (N.). It was
formerly surrounded by the town of Hamilton. (N. S. A.) It is of course
possible that this motte may be much older than the Hamiltons, as the
site of an originally royal castle.

HAY.--First appears in the 12th century, as butler to Malcolm IV. The
family first settled in Lothian, where they had lands at Lochorworth.
The Borthwick family, who got this estate by marriage, obtained a
license from James I. about 1430 to build a castle "on the mote of
Locherwart," and to this castle they gave their own name. (N. S. A.)
No doubt it was the original motte of the Hays. King William gave the
Hays the manor of Errol, in Perthshire, which was made into a barony.
Here is or was the mote of Errol, "a round artificial mound about 20
feet high, and 30 feet in diameter at the top; the platform at the top
surrounded with a low turf wall, and the whole enclosed with a turf
wall at the base, in the form of an equilateral triangle." (N. S. A.;
evidently a triangular bailey.) It is called the Law Knoll, and is
spoken of as a _fortalicium_ in 1546. (N.)

LENNOX.--The earls of Lennox are descended from Arkel, an Englishman,
who received from Malcolm Canmore lands in Dumbartonshire. At Catter,
near the Earl's castle, is a large artificial mound.[1025]

LOCKHART.--Stevenston, in Ayrshire, takes its name from Stephen
Loccard, and Symington, in Lanark, from his son (?), Simon Loccard.
At Stevenson there was formerly a castle, and there still (1845) is
a Castle Hill. Stevenston was given by Richard Morville to Stephen
Loccard about 1170. (N. S. A.) At Symington there was formerly a round
mound, called Law Hill, at the foot of the village, but it has been
levelled. (N. S. A.)

LOGAN.--A Robert Logan witnesses a charter of William the Lion, and
appears later as Dominus Robertus de Logan. The name Robert shows his
Norman origin. At Drumore, near Logan (parish of Kirkmaiden, Wigton),
there was a castle, and there is still a court hill or mote.[1026]
Another mote, at Myroch, in the same parish, is mentioned by Mr Neilson
as the site of the chief messuage of the barony of Logan.

LOVEL.--Settled at Hawick, Roxburghshire. The mote of Hawick, from the
picture in Scott's _Border Antiquities_, seems to be a particularly
fine one. Hawick was a barony, and Le Moit is mentioned in 1511. (N.)

LYLE, or LISLE.--The castle of this Norman family was at Duchal,
Renfrewshire. The plan is clearly that of a motte and bailey, but the
motte is of natural rock.[1027]

MALE, now MELVILLE.--Settled in Haddingtonshire under David I., and
called their seat Melville. Melville Castle is modern. They afterwards
obtained by marriage lands on the Bervie River, in the Mearns. Dr
Christison's map shows a motte near the mouth of the Bervie.

MAXWELL.--Maccus, son of Unwin[1028] (evidently of Scandinavian
origin), received lands on the Tweed from David I., and called his seat
Maccusville, corrupted into Maxwell. There is a motte at Maxwell, near
Kelso. (N.) Maxton, in Roxburghshire, takes its name from him, and
there is a motte called Ringley Hall, on the Tweed, in this parish. (C.
and N. S. A.)

MONTALT, or MOWAT.--Robert de Montalto (Mold, in Flintshire) witnesses
a charter of David I. The family settled in Cromarty. Le Mote at
Cromarty is mentioned in 1470. (N.)

MONTGOMERY.--This family is undoubtedly descended from some one of the
sons of the great Earl Roger of Shrewsbury, settled in Scotland after
the ruin of his family in England. Robert de Montgomerie received
the manor of Eaglesham, Renfrew, from Fitz Alan, the High Steward of
Scotland. The principal messuage of this manor was at Polnoon, 1/2
mile S.E. of Eaglesham. Here Sir John Montgomerie built the castle of
Polnoon about 1388. (N. S. A.) The O.M. seems to show that the ruins
of this castle stand on a motte, probably the original castle of
Montgomerie.

MORVILLE.--Hugh de Morville was a Northamptonshire baron, the life-long
friend of David I.[1029] He founded one of the most powerful families
in the south of Scotland, though after three generations their lands
passed to heiresses, and their chief seat is not even known by name.
But Mr Neilson states that Darnhall, in Peebles, was the head of their
"Black Barony," and that there is a motte there. As Hugh de Morville
gave the church of Borgue to Dryburgh Abbey about 1150, it is probable
that the motte at Boreland of Borgue was one of his castles. The
barony of Beith, in Ayr, given by Richard de Morville to the Abbey of
Kilwinning, has also a motte, which may be reckoned to be the site of
a De Morville castle. Largs, in Ayr, belonged to the De Morvilles, and
has a Castle Hill near the village, which appears to be a motte. (G.)

MOWBRAY.--This well-known Norman family also sent a branch to Scotland.
Amongst other places, about which we have no details, they held
Eckford, in Roxburghshire. In this parish, near the ancient mansion,
is an artificial mount called Haughhead Kipp. (N. S. A.) This seems a
possible motte, but its features are not described.

MURRAY.--Freskin the Fleming came to Scotland under David I., and
received from that king lands in Moray. He built himself a castle
at Duffus, in Elgin, which is on the motte-and-bailey plan.[1030]
The stone keep now on the motte appears to be of the 14th century.
Freskin's posterity took the name of De Moravia, or Moray. (Fig. 44.)

OLIPHANT, or OLIFARD.--Cambuslang, in Lanark, belonged to Walter
Olifard, Justiciary of Lothian in the time of Alexander II. About a
mile E. of the church is a circular mound 20 feet high. It was here
that the Oliphants' castle of Drumsagard formerly stood. (N. S. A.)
Drumsagard was a barony. (N.)

DE QUINCY.--Obtained from William the Lion the manors of Travernant, in
East Lothian, and Leuchars, in Fife. Near the village of Leuchars is
a motte with some slight remains of a stone keep, a deep well in the
centre, and an entrenched bailey, known as the site of the castle of
Leuchars.[1031]

ROSS.--Godfrey de Ros, a vassal of Richard de Morville, held of him the
lands of Stewarton, in Ayr. The _caput_ of the lordship was Castletown,
where Le Mote is spoken of in 1451 (N. and C.). The De Ros were also
the first lords of the barony of Sanquhar. A little lower down the
river Nith than the later castle of Sanquhar is a mote called Ryehill,
and a place anciently manorial. (N.)

SOMERVILLE.--William de Somerville was a Norman to whom David I.
gave the manor of Carnwath, in Lanarkshire. There is a very perfect
entrenched motte at Carnwath (N. S. A. and O.M.), and Le Moit de
Carnwath is mentioned in 1599. (N.)

DE SOULIS.--Followed David I. from Northamptonshire into Scotland, and
received Liddesdale, in Roxburghshire, from him. The motte and bailey
of his original castle still remain, very near the more celebrated but
much later Hermitage Castle.[1032] (Fig. 44.)

VALOIGNES.--Philip de Valoignes and his son William were each
successively chamberlains of Scotland.[1033] One of their estates was
Easter Kilbride, in Lanarkshire, where they had a castle. In this
parish is an artificial mount of earth, with an oval area on top, about
1/4 mile from the present house of Torrance. (N. S. A.)

VAUX, or DE VALLIBUS.--Settled in Scotland under William the Lion. Held
the manors of Dirleton and Golyn, in East Lothian. Dirleton has been
transformed into an Edwardian castle, but from the pictures it appears
to stand on a natural motte of rock. But about 3 miles from Dirleton
the O.M. shows a large motte called Castle Hill, which may possibly be
the original castle of the De Vaux.

WALLACE, or WALLENSIS.--Richard Walensis was the first of this family,
and acquired lands in Ayrshire in David I.'s time. He named his seat
Riccardton, after himself, and the remains of his motte are still
there, a small oval motte called Castle Hill, on which the church of
Riccarton now stands, but which is recognised as having been a "mote
hill." (G.)

To this list must be added a number of royal castles known to have been
built in the 12th century, which, as they were built on mottes, must in
the first instance have been wooden castles.

BANFF.--It seems clear that Banff Castle had a motte, because the
doggerel rhymes of Arthur Johnstone in 1642 say:

    A place was near which was a field until
    Our ancestors did raise it to a hill;
    A stately castle also on it stood.

The _Gazetteer_ says: "The citadel occupied a mount, originally at the
end though now near the middle of the town." The site is still called
Castle Hill. (N. S. A.)

CRAIL, Fife.--The O.M. does not show a motte here. The N. S. A. says
"there was a royal residence here, upon an eminence overlooking the
harbour." That this "eminence" was a motte seems clear from the
_Register of the Great Seal_, quoted by Mr Neilson, which speaks of "Le
Moitt olim Castrum" in 1573.

CUPAR.--There seem to be two mottes here, both raised on a natural
"esker"; the one formerly called the Castle Hill is now called the
School Hill, the school having been built upon it. The other and higher
hill is called the Moot Hill, and is said to be the place where the
earls of Fife used to dispense justice. (N. S. A.) Mr Neilson states
that both are mentioned in the _Registrum_.

DUMFRIES.--Here there were two mottes, one being now the site of a
church, the other, called Castle Dykes, a short distance S. of the
town, on the opposite side of the river. Both no doubt were royal
castles, and Mr Neilson has suggested that as an _old castlestead_
is spoken of in a charter of William the Lion, it implies that a new
castle had recently been built, possibly after the great destruction of
the royal castles in Galloway in 1174.[1034] The Castle Dykes appears
to be the later castle, as it is spoken of in the 16th century. (N.)

DUNSKEATH, Cromarty.--Built by William the Lion in 1179. The castle is
built on a small _moat_ overhanging the sea. (G.)

ELGIN.--Built by William the Lion on a small green hill called Lady
Hill, with conical and precipitous sides. (N. S. A. and G.)

FORFAR.--"The castle stood on a round hill to the N. of the town, and
must have been surrounded by water." (N. S. A.) It was destroyed in
1307. It is called Gallow Hill in the O.M., and is now occupied by
gasworks.

FORRES.--The plan in Chalmers' _Caledonia_ clearly shows a motte, to
which the town appears to have formed a bailey.

INVERNESS.--Built by David I. when he annexed Moray. The site is now
occupied by a gaol, but the O.M. shows it to have been a motte, which
is clearly depicted in old engravings.

INNERMESSAN.--As the lands here appear to have been royal property as
late as the time of David II., the large round motte here may have been
an early royal castle, a conjecture which finds some confirmation in
the name "Boreland of Kingston," which Pont places in the same parish.
(N. S. A.)

JEDBURGH.--Probably built by David I. The site, which is still
called Castle Hill, has been levelled and completely obliterated by
the building of a gaol. Yet an old plan of the town in 1762, in the
possession of the late Mr Laidlaw of Jedburgh, shows the outline of the
castle to have been exactly that of a motte and bailey, though, as no
hachures are given, it is not absolutely convincing.

KINCLEVEN, Perth.--The O.M. shows no earthworks connected with the
present castle, but on the opposite side of the river it places a motte
called Castle Hill, which may very likely be the site of the original
castle.

KIRKCUDBRIGHT.--Dr Christison marks a motte here, to the W. of the
town. The place is called Castle Dykes. Mr Coles says it has an oblong
central mound and a much larger entrenched area.[1035]

LANARK.--Ascribed traditionally to David I. "On a small artificially
shaped hill between the town and the river, at the foot of the street
called Castle Gate, and still bearing the name of Castle Hill, there
stood in former times beyond all doubt a royal castle." (N. S. A.) Mr
Neilson says, "It certainly bears out its reputation as an artificial
mound."

ROSEMARKIE, Cromarty.--Was made a royal burgh by Alexander II., so the
castle must have been originally royal. "Immediately above the town is
a mound of nearly circular form, and level on the top, which seems to
be artificial, and has always been called the Court Hill." (N. S. A.)

Even if we had no other evidence that motte-castles were of Norman
construction, this list would be very significant. But taken in
connection with the evidence for the Norman origin of the English,
Welsh, and Irish mottes, it supplies ample proof that in Scotland, as
elsewhere, the Norman and feudal settlement had its material guarantees
in the castles which were planted all over the land, and that these
castles were the simple structures of earth and wood, whose earthen
remains have been the cause of so much mystification.




CHAPTER XI

MOTTE-CASTLES IN IRELAND


In the year 1169, when the first Norman invaders landed in Ireland,
the private castle had been in existence in England for more than a
hundred years, and had it been suited to the social organisation of
the Irish people, there had been plenty of time for its introduction
into Ireland. Nor are we in a position to deny that some chieftain
with a leaning towards foreign fashions _may_ have built for himself a
castle in the Anglo-Norman style; all we can say is that there is not
the slightest evidence of such a thing.[1036] We have two contemporary
accounts of the Norman settlement in Ireland, the one given by Giraldus
in his _Expugnatio Hibernica_, and the Anglo-Norman poem, edited
by Mr Goddard H. Orpen, under the title of the "Song of Dermot and
the Earl."[1037] Now Giraldus expressly tells us that the Irish did
not use castles, but preferred to take refuge in their forests and
bogs.[1038] The statement is a remarkable one, since Ireland abounds
with defensive works of a very ancient character; are we to suppose
that these were only used in the prehistoric period? But if castles of
the Norman kind had been in general use in Ireland in the 12th century,
we should certainly hear of their having been a serious hindrance to
the invaders. The history of the invasion, however, completely confirms
the statement of Giraldus; we never once hear of the Irish defending
themselves in a castle. When they do stand a siege, it is in a walled
town, and a town which has been walled, not by themselves, but by the
Danes, to whom Giraldus expressly attributes these walls. Moreover,
the repeated insistence of Giraldus on the necessity of systematic
incastellation of the whole country[1039] is proof enough that no such
incastellation existed.

It is true that in some of the earliest Irish literature we hear of
the _dun_, _lis_, or _rath_ (the words are interchangeable), which
encircled the chieftain's house.

Many descriptions of royal abodes in Irish poems are evidently purely
fanciful, but underneath the poetical adornments we can discern the
features of the great wooden hall which appears to have been the
residence of the tribal chieftain, whether Keltic, Norse, or Saxon,
throughout the whole north of Europe in early times.[1040] The
thousands of earthen rings, generally called _raths_, which are still
scattered over Ireland, are believed to be the enclosures of these
kings' or chieftains' homesteads. Were they intended for serious
military defence? We are not in a position to answer this question
categorically, but the plans of a number of them which we have examined
do not suggest anything but a very slight fortification, sufficient to
keep off wolves. At all events we never hear of these raths or duns
standing a siege; the conquering raider comes, sees, and burns.[1041]
We are therefore justified in concluding that they did not at all
correspond to what we mean by a private castle. And most certainly
the motte-castle, with its very small citadel, and its limited
accommodation for the flocks and herds of a tribe, was utterly unsuited
to the requirements of the tribal system.

A good deal of light is thrown on the way in which Irish chieftains
regarded private castles at the time of the invasion by the well-known
story of one who refused a castle offered him by the invaders, saying
that he preferred a castle of bones to a castle of stones. Whether
legendary or not, it represents the natural feeling of a man who had
been accustomed to sleep trustfully in the midst of men of his own
blood, tied to him by the bonds of the clan. The clan system in
Ireland undoubtedly led to great misery through the absence of a
central authority to check the raids of one clan upon another; but
though we occasionally hear of a chieftain being murdered "by his own,"
we have no reason to think that clan loyalty was not sufficient, as
a rule, for the internal safety of the community. So that a popular
chieftain might well refuse a fortification which had every mark of a
hateful and suspicious invader.[1042]

Unfortunately there is--or has been until quite recently--a strong
prejudice in the minds of Irish antiquaries that works of the
motte-and-bailey kind belong to the prehistoric age of Ireland. Irish
scholars indeed admit that the word _mota_ is not found in any Irish
MS. which dates from before the Norman invasion of Ireland.[1043] We
must therefore bear in mind that when they tell us that such and such
an ancient book mentions the "mote" at Naas or elsewhere, what they
mean is that it mentions a _dun_, or _rath_, or _longport_, which
they imagine to be the same as a motte. But this is begging the whole
question. There is not the slightest proof that any of these words
meant a motte. _Dun_ is often taken to mean a hill (perhaps from its
resemblance to Anglo-Saxon _dun_), but Keltic scholars are now agreed
that it is cognate with the German _zaun_ and Anglo-Saxon _tun_,
meaning a fenced enclosure.[1044] It may be applied to a fort on a
hill, but it may equally well be applied to a fort on the flat. _Rath_
is translated _fossa_ in the _Book of Armagh_; Jocelin of Furness
equates it with _murus_.[1045] The rath of Armagh was evidently a
very large enclosure in 1166, containing several streets, houses, and
churches, so it was certainly not a motte.[1046] It is of course not
impossible that the Normans may sometimes have occupied an ancient
fortified site, but we may be sure from the considerations already
urged that the fortifications which they erected were of a wholly
different character to the previous ones, even if they utilised a
portion for their bailey.

It is of course difficult to decide in some cases (both in Ireland and
elsewhere) whether a mound which stands alone without a bailey is a
sepulchral tumulus or a motte. There are some mottes in England and
Scotland which have no baileys attached to them, and do not appear ever
to have had any. In Ireland, the country of magnificent sepulchral
tumuli, it is not wonderful that the barrow and the motte have become
confused in popular language. It would appear, too, that there exist
in Ireland several instances of artificial tumuli which were used for
the inauguration of Irish chieftains, and these have occasionally been
mistaken for mottes.[1047] As Mr Orpen has shown, there are generally
indications in the unsuitability of the sites, in the absence of
real fortification, or in the presence of sepulchral signs, to show
that these tumuli did not belong to the motte class. Magh Adair, for
example, which has been adduced as a motte outside the Norman boundary,
is shown by Mr Orpen to be of quite a different character.

At many sites in Ireland where the Normans are known to have built
castles at an early period of the invasion there are no mottes to be
seen now. It is probable that where the Norman conquerors had both
money and time at their disposal they built stone keeps from the first,
and that the motte-castles, with their wooden towers or _bretasches_,
were built in the times of stress, or were the residences of the less
wealthy under-tenants. But we know from documents that even in John's
reign the important royal castle of Roscrea was built with a motte and
bretasche,[1048] which proves that this type of castle was still so
much esteemed that we may feel reasonably certain that when Giraldus
speaks of "slender defences of turf and stakes" he does not mean
motte-castles, but mere embankments and palisades.[1049]

But there is another reason for the absence of mottes from some of the
early Norman castle sites. Those who have examined the castles of Wales
know that it is rare to find a motte in a castle which has undergone
the complete metamorphoses of the Edwardian[1050] period. These new
castles had no keeps, and necessitated an entire change of plan, which
led either to the destruction of the motte or the building of an
entirely new castle on a different site. The removal of a motte is only
a question of spade labour, and many sites in England can be pointed
out where mottes are known to have existed formerly, but where now not
a vestige is left.[1051] There are many other cases where the Edwardian
castle shows not a trace of any former earthworks, but where a motte
and bailey a little distance off probably represents the original
wooden castle.[1052]

The passion for identifying existing earthworks with sites mentioned
in ancient Irish history or legend has been a most serious hindrance
to the progress of real archæological knowledge in Ireland. It is not
until one begins to look into this matter that one finds out what giddy
guesswork most of these identifications of Irish place-names really
are. O'Donovan was undoubtedly a great Irish scholar, and his editions
of the _Book of Rights_ and the _Annals of the Four Masters_ are of
the highest importance. The topographical notes to these works are
generally accepted as final. But let us see what his method was in this
part of his labours. In the _Book of Rights_, he says very naïvely,
about a place called Ladhrann or Ardladhrann, "I cannot find any place
in Wexford according with the notices of this place except Ardamine,
on the sea-coast, where there is a remarkable _moat_."[1053] No modern
philologist, we think, would admit that Ardamine could be descended
from Ardladhrann. In the same way O'Donovan guessed Treada-na-righ,
"the triple-fossed fort of the kings," to be the motte of Kilfinnane,
near Kilmallock. But this was a pure guess, as he had previously
guessed it to be "one of the forts called Dun-g-Claire." To the
antiquaries of that day one earthwork seemed as good as another, and
differences of type were not considered important.[1054]

       *       *       *       *       *

The following list of early Norman castles in Ireland was first
published in the _Antiquary_ for 1906. It is an attempt to form a
complete list from contemporary historians only, that is, from Giraldus
Cambrensis and the "Song of Dermot," and from the documents published
in Sweetman's _Calendar_, of the Norman castles built in Ireland,
up to the end of John's reign.[1055] Since then, the task has been
taken up on a far more philosophical plan by Mr Goddard H. Orpen,
whose exceptional knowledge of the history of the invasion and the
families of the conquerors has enabled him to trace their settlements
in Ireland as they have never been traced before.[1056] Nevertheless,
it still seems worth while to republish this list, as though within
a limited compass, consistent with the writer's limited knowledge,
it furnishes an adequate test of the correctness of the Norman
theory, on a perfectly sound basis. The list has now the advantage
of being corrected from Mr Orpen's papers, and of being enlarged by
identifications which he has been able to make.[1057]

*ANTRIM[1058] (_Cal._, i., 88).--A royal castle in 1251. Present castle
modern; close to it is a large motte, marked in 25-inch O.M.

AQ'I (_Cal._, i., 13).--Unidentified; perhaps an _alias_ for one of the
Limerick castles, as it was certainly in the county of Limerick.

ARDFINNAN, Tipperary (_Gir._, v., 386).--Built in 1185, immediately
after John's coming to Ireland. No motte; castle is late Edwardian and
partly converted into a modern house; one round tower has ogee windows.
[B. T. S.]

ARDMAYLE, or ARMOLEN, Tipperary (_Cal._, i., 81).--A castle of Theobald
Walter. A motte with half-moon bailey, and earthen wing walls running
up its sides, exactly as stone walls do in later Norman castles. Ruins
of a Perpendicular mansion close to it, and also a square tower with
ogee windows. [B. T. S.] Fig. 45.

ARDNURCHER, or HORSELEAP, King's Co. (_Song of Dermot_ and _Cal._, i.,
145).--A castle of Meiler Fitz Henry's, built in 1192.[1059] An oblong
motte with one certain bailey, and perhaps a second. No masonry but the
remains of a wall or bridge across the fosse. [B. T. S.]

ARDREE, Kildare (_Gir._, v., 356, and _Song_).--The castle built
by Hugh de Lacy for Thomas the Fleming in 1182, was at Ardri, on
the Barrow. There is an artificial mound at Ardree, turned into a
graveyard, and near it a levelled platform above the river, on which
stands Ardree House.[1060] On the west bank of the Barrow, opposite
Ardree, is a low circular motte with ditch and bank, but no bailey. A
piece of Norman pottery with green glaze was found by Mr Stallybrass,
one foot below the surface in the counterscarp bank. Mr Orpen thinks
this motte may have been the castle of Robert de Bigarz, also mentioned
by Giraldus as near Ardree, on the opposite side of the Barrow.

ASKEATON, or HINNESKESTI, Limerick.--Built in 1199, probably by Hamo de
Valoignes.[1061] An excellent instance of a motte-and-bailey castle,
where the motte is of natural rock. The splendid keep and hall are of
the 15th century, but there are two older towers, which might date from
1199. This natural motte has been identified with the ancient Irish
fort of Gephthine (Askeaton = Eas Gephthine), mentioned in the _Book
of Rights_. But this work does not mention any _fort_ at Gephthine,
only the place, in a list which is clearly one of lands (perhaps mensal
lands), not of forts, as it contains many names of plains, and of
tribes, as well as the three isles of Arran.[1062]

*ASKELON, or ESCLUEN (_Cal._, i., 91).--Castle _restored to_ Richard de
Burgh in 1215; the site is placed by Mr Orpen at Carrigogunell, which
is in the parish of Kilkeedy, Limerick.[1063] Carrigogunell has the
ruins of a castle on a natural motte of rock.

*ATHLONE, Roscommon (_Cal._, i., 80).--Built in 1210 by the Justiciar,
John de Gray. The keep is placed on a lofty motte, which has been
revetted with masonry. Turlough O'Connor built a _caislen_ at Athlone
in 1129, but it was not even on the site of the Norman castle, for
which John obtained land from the church, as already stated.

BAGINBUN (_Gir._, i., 13; _Song_, 1406).--Mr Orpen has proved that
this was the spot where Raymond le Gros landed and entrenched himself
for four months.[1064] It is a headland on the sea-coast, and headland
castles seldom have mottes, as they were not needed on a promontory
washed on three sides by the sea. Moreover, Baginbun was of the nature
of a temporary fort rather than a residential castle, and it is to
be noted that Giraldus calls it "a poor sort of a castle of stakes
and sods." Still, the small inner area, ditched off with a double
ditch, and the large area, also ditched, roughly correspond to the
motte-and-bailey plan. [B. T. S.]

BALIMORE EUSTACE, Kildare (_Cal._, i., 28).--A castle of the Archbishop
of Dublin. A motte, with a remarkable platform attached to one side
(_cf._ Wigmore Castle). No bailey now; no stone castle. [B. T. S.]

CAHERCONLISH (Karkinlis, Kakaulis, _Cal._, i., 81).--Castle of Theobald
Fitz Walter. There is nothing left above ground but a chimney of
late date. A few yards from it is a hillock, which has very much the
appearance of a mutilated motte. [E. S. A.] Mr Orpen, however, thinks
that Theobald's castle may have been at Knockatancashlane, "the hill of
the old castle," a townland a little to the north of Caherconlish.[1065]

CARBURY, Kildare.--The _Song_ says Meiler Fitz Henry first got
Carbury, so the castle was probably his. It is a motte with two
baileys, one of imperfect outline, the other a curious little
half-circle. A 15th-century castle is built against the side of the
motte. [B. T. S.]

CARLINGFORD, Louth (_Cal._, i., 95).--Apparently a royal castle
(_Cal._, i., 156), first mentioned in 1215. It stands on a rock, which
might possibly have been a former motte. There certainly has been a
former castle, for the present ruin is Edwardian in plan and in every
detail. [E. S. A.]

CARRICK, Wexford (_Gir._, v., 245).--This again seems to be one of
the temporary forts built by the first invaders (in this case Fitz
Stephen), in a strong natural situation, and Giraldus applies to it the
same contemptuous language as to Baginbun. There is no motte, but an
oval area of 45 yards by 25 is ditched and banked; a modern imitation
of a round tower stands within the enclosure. [B. T. S.]

CARRICKFERGUS, Antrim (_Cal._, i., 107).--This was probably one of
the castles built by John de Courcy, the conqueror of Ulster. The
gatehouse and mural towers are late, but the keep may well be of De
Courcy's time, and furnishes an excellent instance of a castle on the
keep-and-bailey plan, built by the Normans in stone from the beginning.
[E. S. A.]

CASTLETOWN DELVIN, Westmeath (_Gir._, v., 356).--Castle of Gilbert
de Nungent. A motte, with a garden at base, which may have been the
bailey; near it the stone castle, a keep with round towers at the
angles, probably not as early as John's reign. [B. T. S.]

CLONARD, Meath (_Gir._, v., 356).--Built by Hugh de Lacy about 1182. A
motte, with broad ditch and curious little oblong bailey; no remains in
masonry. [B. T. S.]

CLONMACNOISE, King's Co. (_Cal._, i., 94).--First contemporary mention
1215; the _Annals of Loch Cè_ say it was built in 1214 "by the
foreigners." A royal castle. A large motte with bailey attached; the
wing walls of the bailey run up the motte. The importance of the castle
is shown by the fact that a stone keep was added not very long after it
was built. [B. T. S.]

*COLLACHT (_Gir._, v., 355).--Castle of John of Hereford. Collacht
appears to be a scribal error for Tullaght, now Tullow, Carlow.[1066]
The site of the castle is marked on the 6-inch O.M.; it has been
visited by Mr G. H. Orpen, who found very clear indications of a motte
and bailey. (See Appendix L.)

CROMETH (_Cal._, i., 91).--Castle of Maurice FitzGerald. Supposed
to be Croom, Limerick, though the identification is by no means
certain.[1067] There are the ruins of an Edwardian castle at Croom; no
motte. [E. S. A.]

DOWNPATRICK, Down (_Gir._, v., 345).--The traveller approaching
Downpatrick sees a number of small hills which no doubt have once
been islands rising out of the swamps of the Quoyle. On one of these
hills stands the town and its cathedral; on another, to the east, but
separated from the town by a very steep descent and a brook, stands
a motte and bailey of the usual Norman type. It occupies the whole
summit of the small hill, so that the banks of the bailey are at a
great height above the outer ditch, which is carried round the base of
the hill (compare Skipsea). The motte, which is not a very large one,
has had an earthen breastwork round the top, now much broken away.
Its ditch falls into the ditch of the bailey, but at a higher level.
The bailey is semilunar, extending round about three-quarters of the
circumference of the motte. There is not the slightest sign of masonry.
As the size of this work has been greatly exaggerated, it is as well to
say that when measured on the 25-inch O.M. with a planimeter, its area
proves to be 3.9 acres; the area of the motte and its ditch .9, leaving
3 acres for the bailey. [E. S. A.] Fig. 45.

This thoroughly Norman-French castle, which was formerly called a
Danish fort, has lately been baptised as Rathceltchair, and supposed
to be the work of a mythical hero of the 1st century A.D. Mr Orpen,
however, has disposed of this fancy by showing that the name
Rathceltchair belonged in pre-Norman times to the enclosure of the
ancient church and monastery which stood on the _other_ hill.[1068]
We may therefore unhesitatingly ascribe this motte-castle to John de
Courcy, who first put up a slender fortification within the town walls
to defend himself against temporary attack,[1069] but afterwards built
a regular castle, for which this island offered a most favourable
site.[1070] A stone castle was built inside the town at a later period;
it is now entirely destroyed.

DROGHEDA, Louth (_Cal._, i., 93).--First mention 1203, but Mr Orpen
thinks it probable that it was one of the castles built by Hugh de
Lacy, who died in 1186. A high motte, with a round and a square bailey,
just outside the town walls;[1071] called the Mill Mount in the time
of Cromwell, who occupied it; he mentions that it had a good ditch,
strongly palisadoed.[1072] No stone castle, though much of the
bailey wall remains; a late martello tower on top of motte. [B. T. S.]
Fig. 45.

[Illustration: FIG. 45.--IRISH MOTTE-CASTLES. Ardmayle. Downpatrick.
Drogheda. Castlenock.]

DULEEK, Meath (the castrum Duvelescense of _Giraldus_, v.,
313).--Probably first built by Hugh de Lacy; restored by Raymond le
Gros in 1173. The motte is destroyed, but an old weaver living in the
village in 1906 says that it existed in the time of his father, who
used to roll stones down it in his youth. It was in the angle between
two streams, and there is still a slight trace of it. No stone castle.
[B. T. S.]

DUNAMASE, Queen's Co. (Dumath, _Cal._, i., 100).--First mentioned in
1215 as a castle of William Marshall's, which makes it not unlikely
that it was originally built by Strongbow. The plan of this castle is
the motte-and-bailey plan, but the place of the motte is taken by a
natural rock, isolated by a ditch. There are three baileys, descending
the hill. The stone keep on the summit is of the 15th or 16th century.
[B. T. S.]

DUNGARVAN, Waterford (_Cal._, i., 89).--Granted to Thomas Fitz Antony
in 1215. To the west of the town is a motte called Gallowshill; it has
no bailey, but some trace of a circumvallation. The castle east of the
river is not earlier than the 14th or 15th century. [B. T. S.]

*DURROW, King's Co. (_Gir._, v., 387).--A castle of Hugh de Lacy's;
he was murdered while he was building it, because he had chosen the
enclosure of the church for his bailey.[1073] A plan in _Journ. R. S.
A. I._, xxix., 227, shows clearly the motte and bailey, though the
writer mistakes for separate mounds what are clearly broken portions of
the vallum. It is possible that the bailey may have followed the line
of the ancient _rath_ of the church, but it would almost certainly be a
much stronger affair.

*FAVORIE = FORE, Westmeath.--I owe this identification to Mr Orpen. As
Hugh de Lacy founded or endowed the monastery at Fore,[1074] this was
probably one of his castles, but the first mention is in 1215 (_Cal._,
i., 95). Mr Westropp mentions the oval motte of Fore with its bailey in
his list of "complex motes."[1075]

FERNS, Wexford (_Gir._, v., 326).--A castle was built by Walter the
German _near_ Ferns. Ferns is spoken of as a city in the time of King
Dermot. There is no motte at Ferns; the stone castle has a keep, which
is certainly not earlier than the time of Henry III. [B. T. S.]

*FOTHERET ONOLAN, castle of Raymond le Gros (_Gir._, v., 355).--Mr
Orpen identifies this with Castlemore, near Tullow, Co. Carlow.
There is an oval motte, and a rectangular bailey with indications of
masonry.[1076]

GALTRIM, Meath.--Identified by Mr Orpen with the castle of Hugh de
Hose, or Hussey, mentioned in the "Song of Dermot." Destroyed in 1176;
no stone castle. An oval motte; bailey indistinctly traceable. [B. T.
S.]

GEASHILL, King's Co. (_Cal._, i., 30).--Mentioned in 1203 as a castle
of William, Earl Marshall. There are remains of a motte, on which
stands a 14th-century keep; but the whole site has been so pulled about
in making a modern house, drive, and gardens, that nothing more can be
made of the plan. The motte, however, is plain, though mutilated. [E.
S. A.]

GRANARD, Longford (_Cal._, i., 95).--Built by Richard Tuit in
1199.[1077] A magnificent motte, with a very wide ditch, and a small
fan-shaped bailey. Foundations of a shell wall round the top of the
motte, and of a small round tower in the centre. [B. T. S.]

*HINCHELEDER, or INCHELEFYRE (_Cal._, i, 95).--Said by Butler (_Notices
of Trim Castle_, 12) to be Inchleffer, Meath, a castle of Hugh de Lacy.
No further information.

JOHN DE CLAHULL'S CASTLE.--Mr Orpen believes this to be Killeshin,
Queen's Co., as it corresponds to the description in the _Song_, "entre
Eboy et Lethelyn." There is a motte there, and traditions of a town.

*KARAKITEL, or CARRICKITTLE, Limerick (_Cal._, i., 14).--Castle of
William de Naas in 1199. There was a remarkable natural motte of rock
here, with the foundations of a castle upon it, now destroyed.[1078]

*KILLAMLUN (_Cal._, i., 53).--Identified by Mr Orpen with Killallon,
Meath, where there is a large motte. There is a stone passage into
this motte, but no evidence has been brought forward to prove that it
is of the same nature as the prehistoric _souterrains_ so common in
Ireland.[1079] In England there is a remarkable instance at Oxford of a
well-chamber built inside a motte.

KILLARE, Westmeath (_Gir._, v., 356).--A castle of Hugh de Lacy,
built in 1184;[1080] burnt in 1187. A good motte, with ditch and
well-preserved bank on counterscarp; no bailey. No stone castle. [B. T.
S.]

KILBIXIE, Westmeath.--Identified by Mr Orpen with Kelbery, given to
Geoffrey de Constantin (_Song_, 3154); the castle is mentioned in a
charter of Walter de Lacy, as well as in the _Annals of Loch Cè_, which
state that it was built in 1192. A motte, with a broad ditch, and no
bailey; but on the W. side the counterscarp bank of the ditch widens
out into a sort of narrow half-moon terrace. This peculiarity may be
noted in several other Irish castles. Foundations of an oblong shell on
top of motte, and of a small square tower in the centre of this ward.
[B. T. S.]

*KILFEAKLE, Tipperary (_Cal._, i., 29).--A castle of William de Burgh.
Built in 1193.[1081] A motte and bailey; trace of a stone wing wall
down the motte.[1082]

*KILMEHAL (_Cal._, i., 44).--Mr Orpen regards the identification of
this castle with Kilmallock as extremely doubtful.

*KILMORE (_Cal._, i., 95).--Restored to Walter de Lacy in 1215.
Identified with Kilmore, near Lough Oughter, Cavan.[1083] Mr Westropp
mentions the motte at this place, which is outside the Anglo-Norman
area. The castle was wrecked in 1225 or 1226, and no more is heard of
it. The Anglo-Norman advance in this direction failed.

*KILSANTAN, Londonderry (_Cal._, i., 70).--Built by John de Courcy in
1197.[1084] Now called Kilsandal, or Mount Sandal, a large motte on the
Bann, not far from Coleraine. The castle of Coleraine, inside the town,
was built in 1214, apparently of stone,[1085] and probably superseded
the castle of Kilsandal.

KILTINAN, Tipperary (_Cal._, i., 94).--Castle of Philip of Worcester
in 1215. No motte; a headland castle overhanging a river valley. The
castle has not only undergone a late Edwardian transformation, but has
been cut up to make a modern mansion and farm buildings. No fosses or
earthworks remain. [E. S. A.]

KNOCK, or CASTLEKNOCK, Dublin (_Cal._, i., 81).--Castle of Hugh Tyrrel.
An oval motte, walled round the top, carrying on its edge a smaller
motte (with traces of a ditch) on which stand the ruins of an octagonal
keep. No other bailey; ditch and bank double for more than half the
circumference. [B. T. S.] Fig. 45.

*KNOCKGRAFFAN, Tipperary (_Cal._, i., 27).--Castle of William de
Braose in 1202. One of the finest mottes to be seen anywhere. Built
in 1192, at the same time as the castle of Kilfeakle.[1086] The motte
is 55 feet high, has a wide ditch and high counterscarp bank, which
is also carried round the ditch of the "hatchet-shaped" bailey, in
proper Norman fashion. "There are indications of a rectangular stone
building on the flat summit of the mote, and there are extensive stone
foundations in the bailey."[1087]

*LAGELACHON (_Cal._, i., 95).--Probably Loughan or Castlekieran, in
which parish is the great motte of Derver.[1088]

LEA, Queen's Co. (_Cal._, i., 30).--Castle of William, Earl Marshall,
in 1203. A motte with two baileys; motte entirely occupied, and partly
mutilated by a 13th-century keep, with two large roundels. [B. T. S.]

LEIGHLIN, Carlow.--Mr Orpen has shown that the fine motte of
Ballyknockan answers to the description given by Giraldus of the site
of the castle of Lechlin built by Hugh de Lacy.[1089] There is a trace
of a possible bailey. The stone castle called Black Castle at Leighlin
Bridge is of very late date. Those who believe that we have authentic
history of Ireland in the 3rd century B.C. will be able to believe
with Dr Joyce that the description of the annalists identifies this
motte with the site of the ancient palace of Dinn Righ, burnt by the
chieftain Maen at that date! [B. T. S.]

LISMORE, Waterford (_Gir._, i., 386).--About a quarter of a mile
from Lismore, above a ford of the river, is an excellent specimen of
a Norman motte and bailey, called the Round Hill. The name of the
prehistoric fort of Dunsginne has lately been applied to it, but purely
by guesswork.[1090] The _Song_ says that Henry II. intended to build a
castle at Lismore, and that it knows not why he put it off. Possibly
he may have placed these earthworks here, and never added the wooden
castle, or else this is the site of the castle which was built by his
son John in 1185. The castle inside the town is certainly later than
the time of John, as although much modernised it is clearly Edwardian
in plan. The Norman fragments incorporated in the walls probably
belonged to the abbey of St Carthagh, on the site of which the town
castle is said to have been built. The so-called King John's Tower is
only a mural tower, not a keep. [B. T. S.]

*LOUTH, or LUVETH (_Cal._, i., 30).--A royal castle in 1204, but
it must have been in existence as early as 1196, when the town and
castle of Louth were burnt by Niall MacMahon.[1091] This was probably
the "Fairy Mount" at Louth, of which a plan is given in Wright's
_Louthiana_. This plan shows "the old town trench," starting from
opposite sides of the motte, so that the castle stood on the line of
the town banks. The motte was ditched and banked round, but the plan
does not show any bailey or any entrance.

*LOSKE (CAL., i., 30).--Mr Orpen has pointed out to the writer that
this cannot be Lusk, which was a castle of the Archbishop of Dublin,
while Loske belonged to Theobald Walter, and is not yet identified.

*LOXHINDY (_Cal._, i., 95).--Mr Orpen identifies this name with
Loughsendy, or Ballymore Loughsendy, Westmeath, where there is a
motte.[1092]

NAAS, Kildare (_Gir._, v., 100).--The _dun_ of Naas is mentioned in the
_Book of Rights_, p. 251, and in the _Tripartite Life of St Patrick_.
By the _Dindsenchas_ it is attributed to the legendary Princess
Tuiltinn in 277 A.D. On this "evidence" the motte at Naas has been
classed as prehistoric. But as we have seen, a _dun_ does not mean a
motte, or even a hill, but an enclosure. Naas was part of the share
which fell to the famous Anglo-Norman leader, Maurice FitzGerald, and
the earthworks are quite of the Norman pattern;[1093] a good motte,
ditched and banked, with trace of a small bailey attached. The terrace
round the flank of the motte may be no older than the modern buildings
on the summit.[1094] [B. T. S.]

NAVAN, Meath.--The _Song_ says Navan was given to Jocelin de Nangle,
and it is known that the castle of the Nangles was at Navan. A lofty
motte, with a very small semilunar platform below, formed by broadening
out a part of the counterscarp bank of the ditch. (Compare Kilbixie.)
[B. T. S.]

NOBBER, Meath (_Cal._, i., 104).--A castle of Hugh de Lacy. A motte,
with traces of a breastwork round the top, and wing banks running down
to what remains of the bailey on the S. Two curious little terraces on
the N. side of the motte. No masonry. [B. T. S.]

RATH' (_Cal._, i., 95).--This castle, evidently one of the most
important in Ulster, but hitherto unidentified, has been shown by Mr
Orpen to be the famous castle of Dundrum, Down.[1095] This castle is
situated on a natural motte of rock, no doubt scarped by art, with a
deep ditch cut through the rock, and a bailey attached. The top of the
motte contains a small ward fortified in stone, and a round keep. It
is very doubtful whether this keep is as old as the time of John de
Courcy, to whom the castle is popularly attributed; for the round keep
without buttresses hardly appears in England before the reign of Henry
III. [E. S. A.]

RATHWIRE, Meath.--Rathwire was the portion of Robert de Lacy (_Song_,
3150), and a castle was built here by Hugh de Lacy.[1096] There is
a motte and bailey, with considerable remains of foundations in the
bailey, and one wing bank going up the motte. [B. T. S.]

*RATOUTH, Meath, now RATOATH (_Cal._, i., 110).--A castle of Hugh de
Lacy. There is "a conspicuous mount" near the church, about which there
is a legend that Malachy, first king of all Ireland, held a convention
of states (Lewis). It is marked in the map.

*ROKEREL (_Cal._, i., 81).--Unidentified.

ROSCREA, Tipperary (_Cal._, i., 81).--A motte and bretasche were
built here in King John's reign, as is recorded in an inquisition of
29 Henry III. (_Cal._, i., 412). There is no motte now at Roscrea,
but an Edwardian castle with mural towers and no keep; a 14th-century
gatehouse tower. Here we have a proved instance of a motte completely
swept away by an Edwardian transformation.[1097] [E. S. A.]

SKREEN, Meath.--Giraldus mentions the castle of Adam de Futepoi, and
as Skreen was his barony, his castle must have been at Skreen. In the
grounds of the modern castellated house at Skreen there is a motte,
11 feet high (probably lowered), with a terrace round its flank; some
slight traces of a bailey. [B. T. S.]

SLANE, Meath.--The _Song_ relates the erection of a motte by Richard
the Fleming: "un mot fist cil jeter pur ses enemis grever."[1098] It
also tells of its destruction by the Irish, but does not give its
name, which is supplied by the _Annals of Ulster_. Probably Richard
the Fleming restored his motte after its destruction, for there is
still a motte on the hill of Slane, with a large annular bailey,[1099]
quite large enough for the "100 foreigners, besides women and children
and horses," who were in it when it was taken. The motte has still a
slight breastwork round the top. The modern castle of the Marquis of
Conyngham, below, incorporates half a round tower of 13th-century work,
belonging no doubt to the stone castle which succeeded the motte.[1100]
[B. T. S.]

THURLES, Tipperary (Dorles, _Cal._, i., 81).--A castle of Theobald
Walter. Thurles Castle has a late keep with trefoil windows, and
according to Grose was built by the Earl of Ormond in 1328. From
information on the spot it appears that there used to be a motte in the
gardens behind the castle; mentioned also by Lewis. [B. T. S.]

TIBRAGHNY, or TIPPERAGHNY, Kilkenny (_Gir._, i., 386; _Cal._, i.,
19).--Granted to William de Burgh in 1200; built by John in 1185.[1101]
A motte, with ditch and bank, and some trace of a half-moon bailey to
the north. About 200 yards away is the stone castle, a late keep with
ogee windows. [B. T. S.]

TIMAHOE, Queen's Co. (_Gir._, i., 356).--Built by Hugh de Lacy for
Meiler Fitz Henry. A motte, called the Rath of Ballynaclogh, half a
mile west of the village. The bailey, the banks and ditches of which
seem remarkably well preserved, is almost circular, but the motte is
placed at its edge, not concentrically. There are wing-banks running up
the motte. Near it are the ruins of a stone castle built in Elizabeth's
reign (Grose). [B. T. S.]

TRIM, Meath.--The _Song_ tells of the erection of this castle by Hugh
de Lacy, and how in his absence the _meysun_ (the keep--doubtless
wooden) was burnt by the Irish, and the _mot_ levelled with the ground.
This express evidence that the first castle at Trim had a motte is
of great value, because there is no motte there now. The castle was
restored by Raymond le Gros,[1102] but so quickly that the present
remarkable keep can hardly have been built at that date.[1103] [B. T.
S.]

*TRISTERDERMOT (_Gir._, v., 356).--Castle of Walter de Riddlesford.
Tristerdermot is now Castledermot; there used to be a _rath_ of some
kind here close to the town. But Mr Orpen inclines to believe that
the castle Giraldus alludes to was at Kilkea, another manor of De
Riddlesford's, where there is a motte, near the modern castle. "In the
early English versions of the _Expugnatio_ Kilcae is put instead of
Tristerdermot as the place where Walter de Riddlesford's castle was
built."[1104]

*TYPERMESAN (_Cal._, i., 110).--Mr Orpen writes that this name occurs
again in a list of churches in the deanery of Fore, which includes
all the parish names in the half barony of Fore, except Oldcastle and
Killeagh. He suspects that Typermesan is now known as Oldcastle, "where
there is a remarkably well-preserved motte and raised bailey."[1105]

WATERFORD (_Cal._, i., 89).--We are not told whether Strongbow built a
castle here when he took the town from the Ostmen in 1170. The castle
is not mentioned till 1215, when it was granted by John to Thomas
Fitz-Antony. Waterford was a walled town in 1170, and had a tower
called Reginald's Tower, which seems to have been the residence of
the two Danish chieftains, as they were taken prisoners there. Here
too, Henry II. imprisoned Fitz Stephen.[1106] It is possible that this
tower, as Mr Orpen supposes,[1107] may have been considered as the
castle of Waterford. But the existing "Ring tower" on the line of the
walls, which is sometimes called Reginald's Tower, is certainly a round
mural tower of the 13th century; there are others of similar masonry on
the walls. [B. T. S.]

*WEXFORD (_Gir._, v., 314).--Probably built by Maurice Prendergast;
first mentioned when taken from his sons in 1176. Mr Orpen writes:
"The site of Wexford Castle is an artificial mound. Two of the scarped
sides still remain, and the other two are built up above streets. When
recently laying some drainpipes, the workmen came upon no rock, but
only made earth."

WICKLOW (_Gir._, i., 298).--Existing when Henry II. left Ireland
in 1173; he gave it to Strongbow. The Black Castle at Wicklow is a
headland castle; it preserves the motte-and-bailey plan, though there
is no motte, as there is a small triangular inner ward (about thirty
paces each side) several feet higher than the outer bailey, from which
it is separated by a very deep ditch cut through the rock. [B. T. S.]

We have here a list of seventy-two castles mentioned in the contemporary
history of the Norman invasion. If the list is reduced by omitting
Aq'i, Kilmehal, Loske, Rokerel, and Incheleder, which are not yet
identified, and five castles of which the identification may be
considered doubtful, Caherconlish, Croom, Clahull's Castle, Lagelachan,
and Typermesan, sixty-two castles are left, and out of these sixty-two,
fifty-two have or had mottes.[1108] In five cases the place of the
motte is taken by a natural rock, helped by art; but as the idea and
plan are the same it is legitimately classed as the same type.

This list might easily have been enlarged by the addition of many
castles mentioned in the various Irish annals as having been built
by the Normans. But this would have involved the identification of
a number of difficult names, a labour to which the writer's limited
knowledge of Irish topography was not equal. The greater number of
these sites have now been identified by Mr Orpen, and to his papers, so
frequently cited above, we must refer the reader who wishes to study
the fullest form of the argument sketched in these pages.

One can easily sympathise with the feelings of those who, having
always looked upon these mottes as monuments of ancient Ireland, are
loath to part with them to the Norman robber. Many of us have had
similar feelings about the mottes of England, some of which we had been
taught to regard as the work of that heroic pair, Edward the Elder and
Ethelfleda. But these feelings evaporated when we came to realise that
it would have been highly unpatriotic in these founders of the British
empire to have built little castles for their own personal safety,
instead of building cities which were "to shelter all the folk," in the
words of Ethelfleda's charter to Worcester. In like manner, wretched as
were the intertribal wars of Ireland, it would have been a disgrace to
the Irish chieftains if they had consulted solely their own defence by
building these little strongholds for their personal use.

The Irish motte-castles furnish us with interesting proof that this
type of castle was commonly used, not only as late as the reign
of Henry II., but also in the reigns of his sons, Richard I. and
John;[1109] that is to say, at a time when castle-building in stone was
receiving remarkable developments at the hands of Richard I. and Philip
Augustus of France. This, however, need not surprise us, since we know
that as late as 1242, Henry III. was building a motte and wooden
castle in the Isle of Rhé, at the mouth of the Garonne.[1110] But those
who imagine that the Normans built stone castles everywhere in England,
Wales, and Ireland, will have to reconsider their views.

_Note._--Mr Orpen's work on _Ireland under the Normans_ did not appear
until too late for use in this chapter. The reader is referred to it
for a more careful tracing of the history and archæology of the Norman
settlements in Ireland.




CHAPTER XII

STONE CASTLES OF THE NORMAN PERIOD


It may be a surprise to some of our readers to learn how very few stone
castles there are in England which can certainly be ascribed to the
first period of the Norman Conquest, that is to the 11th century. When
we have named the Tower of London, Colchester, the recently excavated
foundations of the remarkable keep at Pevensey, and perhaps the ruined
keep of Bramber, we have completed the list, as far as our present
knowledge goes, though possibly future excavations may add a few
others.[1111]

It is obvious that so small a number of instances furnishes a very
slender basis for generalisations as to the characteristics of early
Norman keeps, if we ask in what respect they differed from those of the
12th century. But it is the object of this chapter to suggest research,
rather than to lay down conclusions. The four early instances mentioned
should be compared with the earliest keeps of France, the country where
the pattern was developed. This has not yet been done in any serious
way, nor does the present writer pretend to the knowledge which would
be necessary for such a comparison.[1112] But data exist, which, if
they were used in the right way, would greatly add to our knowledge.

In the first place, we have a list of the castles built by Fulk Nerra,
Count of Anjou, at the end of the 10th and the beginning of the 11th
century, during his life-long struggle with the Counts of Blois for
the possession of Touraine. This list may be regarded as authentic,
as it is given by his grandson, Fulk Rechin, in the remarkable
historical fragment which he has bequeathed to us.[1113] The list is
as follows:--_In Touraine_: Langeais, Chaumont-sur-Loire, Montrésor,
St Maure. _In Poitou_: Mirabeau (N.W. of Poitiers), Montcontour,
Faye-la-Vineuse, Musterolum (Montreuil-Bonnin), Passavent, Maulevrier.
_In Anjou_: Baugé, Chateau-Gontier, Durtal. "Et multa alia," adds
Fulk's grandson. Nine of these others are mentioned by the chroniclers:
Montbazon, Semblançay, Montboyau, St Florent-le-Vieil, Chateaufort
near Langeais, Chérament, Montrevault, Montfaucon, and Mateflon.
Many of these were undoubtedly wooden castles, with wooden keeps on
mottes.[1114] In many other cases the ancient fabric has been replaced
by a building of the Renaissance period. Whether any remains of stone
donjons built by Fulk Nerra exist at any of these places except at
Langeais, the writer has been unable to find out; probably Langeais
is the only one; but French archæologists are agreed that the ruined
tower which stands on the ridge above the 15th-century castle of
Langeais is the work of this count,[1115] a venerable fragment of a
10th-century keep.[1116]

Unfortunately only two sides of this tower and the foundations of
the other sides remain. The walls are only 3 feet 6 inches thick,
contrasting strikingly with the castles of the 12th and 13th centuries,
where the usual thickness is 10 feet, which is often exceeded. This
points to a date before any great improvement had taken place in
assaulting-machinery. The masonry is what French architects call
_petit appareil_, very small stones, but regularly coursed. There is
no herring-bone work. The buttresses, of which there are five on the
front, certainly suggest a later date, from the size of the ashlar with
which they are faced, and from their considerable projection (3 feet on
the entrance wall, 2 on the front). There is no sign of a forebuilding.
There are only two storeys above the basement. The floors have been
supported on ledges, not on vaults. The doorway, a plain round arch,
with bar-holes, is on the first floor;[1117] it is now only a few feet
above the ground, but probably the basement has been partially filled
up with rubbish. The first storey is quite windowless in the walls
which remain. There are no fireplaces nor any loopholes in these two
fragments. In the second storey there are three rather small windows
and one very large one;[1118] they are round arched, have no splay, and
their voussoirs are of narrow stones alternated with tiles. In these
details they resemble the Early Romanesque, which in England we call
Anglo-Saxon.

The Tower of London and Colchester keep are some seventy or eighty
years later than that of Langeais, and if we attempt to compare them,
we must bear in mind that Langeais was the work of a noble who was
always in the throes of an acute struggle with a powerful rival,
whereas the Tower and Colchester Castle were built by a king who had
reached a position of power and wealth beyond that of any neighbouring
sovereign.[1119] Langeais is but a small affair compared with these
other two keeps. The larger area,[1120] thicker walls, the angle towers
with their provision of stairways, the splayed windows [of Colchester],
the fireplaces, the chapels with round apses, the mural gallery [of the
Tower] cannot be definitely pronounced to be instances of development
unless we have other instances than Langeais to compare with them.
De Caumont mentions Chateau du Pin (Calvados), Lithaire (Manche),
Beaugency-sur-Loire, Nogent-le-Rotrou (Eure et Loire), Tour de l'Islot
(Seine et Oise), St Suzanne (Mayenne), and Tour de Broue (Charente
Inf.), as instances of keeps of the 11th century.[1121] These should
be carefully examined by the student of castle architecture, and De
Caumont's statements as to their date should be verified. Not having
had the opportunity of doing this, we will only ask what features the
keeps of Langeais, London, and Colchester have in common, which may
serve as marks of an earlier date than the 12th century.[1122] The
square or oblong form and the entrance on the first floor are common
to all three, but also to the keeps of the first three-quarters of
the 12th century. The absence of a forebuilding is probably an early
sign,[1123] and so is the extensive use of tiles.[1124] The chapel
with a round apse which projects externally only occurs in the keeps
of London and Colchester, and in the ruins of Pevensey keep.[1125] The
absence of a plinth is believed by Enlart to be an early token.[1126]
But Colchester has a plinth and so has the Tower. It is, however,
very possible that in both cases the plinth is a later addition; at
Colchester it is of different stone to the rest of the building, and
may belong to the repairs of Henry II., who was working on this castle
in 1169; while the Tower has undergone so many alterations in the
course of its eight hundred years of existence that it is difficult to
say whether the rudimentary plinth which it still possesses is original
or not.

Wide-jointed masonry is generally recognised by architectural students
as a mark of the early Norman style. Even this is a test which may
sometimes deceive; certain kinds of ashlar are very liable to weather
at the edges, and when the wall has been pointed at a comparatively
recent period, a false appearance of wide joints is produced.
Moreover, there are instances of wide-jointed masonry throughout the
12th century. The use of rubble instead of ashlar is common at all
dates, and depends no doubt on local conditions, the local provision
of stone, or the affluence or poverty of the castle-builder. We are
probably justified in laying down as a general rule that the dimensions
of the ashlar stones increase as the Middle Ages advance. There is a
gradual transition from the _petit appareil_ of Fulk Nerra's castle
to the large blocks of well-set stone which were used in the 15th
century.[1127] But this law is liable to many exceptions, and cannot be
relied upon as a test of date unless other signs are present. The Tower
of London is built of Kentish rag; Colchester keep of small cement
stones (septaria), which whether they are re-used Roman stones or not,
resemble very much in size the masonry of Langeais. It is of course
unnecessary to say to anyone who is in the least acquainted with Norman
architecture that all Norman walls of ashlar are of the core-and-facing
kind, an internal and an external shell of ashlar, filled up with
rubble; a technique which was inherited from Roman times in Gaul, but
which was not followed by the Anglo-Saxons.[1128]

The presence or absence of fireplaces and chimneys is not a test of
date. Colchester is certainly an early keep,[1129] but it is well
provided with fireplaces which appear to be original. These fireplaces
have not proper chimneys, but only holes in the wall a little above the
fireplace. But this rudimentary form of chimney is found as late as
Henry II.'s keep at Orford, and there is said to be documentary mention
of a proper chimney as early as 816 in the monastery of St Gall.[1130]
The entire absence of fireplaces is no proof of early date, for in
Henry II.'s keep at the Peak in Derbyshire, the walls of which are
almost perfect (except for their ashlar coats) there are no fireplaces
at all, nor are there any in the 13th-century keep of Pembroke. It is
possible that in these cases a free standing fireplace in the middle
of the room, with a chimney carried up to the roof, was used. Such a
fireplace is described by the poet, Chrestien of Troyes, but no example
is known to exist.[1131]

But apart from details, if we look at the general plan of these four
early stone castles, we shall see that it is exactly similar. It is
the keep-and-bailey plan, the plan which prevailed from the 10th to
the 13th century, and was not even superseded by the introduction of
the keepless castle in the latter century.[1132] The motte-and-bailey
type was of course only another version of the keep-and-bailey. In
this primitive type of castle the all-important thing was the keep or
donjon.[1133] Besides the donjon there was little else but a rampart
and ditch. "Until the middle of the 12th century, and in the simpler
examples of the epochs which followed, the donjon may be said to
constitute in itself the whole castle."[1134] Piper states that up
to the time of the Crusades German castles do not seem to have been
furnished with mural towers.[1135] Köhler, whose work treats of French
and English castles as well as German, says that mural towers did
not become general till the second half of the 12th century.[1136]
Nevertheless, as it is highly probable that the baileys of castles were
defended at first with only wooden ramparts on earthen banks, even
when the donjon was of stone, it is not unlikely that mural towers of
wood may have existed at an earlier period than these writers suppose.
It is, however, in favour of the general absence of mural towers that
the word _turris_, even in 12th-century records, invariably means the
_keep_, as though no other towers existed.[1137]

That the baileys of some of the most important castles in England
had only these wooden and earthen defences, even as late as the 13th
century, can be amply proved from the _Close Rolls_.[1138] Colchester
Castle had only a timber wall on the banks of its bailey as late as
1215, and in 1219 this _palicium_ was blown down and an order issued
for its reconstruction.[1139]

The arrangements in the stone donjons were probably the same as those
we have already described when writing of the wooden ones.[1140] The
basement was the storehouse for provisions,[1141] the first floor was
generally the guardhouse, the second the habitation, of the lord and
lady. Where there were three or four storeys, the arrangements varied,
and the finest rooms are often found on the third floor. An oratory was
probably an invariable feature, though it cannot always be detected in
ruined keeps. One of Mr Clark's most pronounced mistakes was his idea
that these keeps were merely towers of refuge used only in time of
war.[1142] History abounds with evidence that they were the permanent
residences of the nobles of the 11th and 12th centuries. The cooking,
as a rule, was carried on in a separate building, of which there are
remains in some places.[1143]

Occasionally we find a variant of the keep-and-bailey type, which
we may call the gatehouse keep. The most remarkable instance of
this kind in England is Exeter, which appears never to have had any
keep but the primitive gatehouse, undoubtedly the work of Baldwin de
Moeles, the first builder of the castle. In Normandy, De Caumont gives
several instances of gatehouse keeps. Plessis-Grimoult (which has been
visited by the writer) has a fragment of a gatehouse tower, but has
also a mural tower on the line of the walls; as the castle was ruined
and abandoned in 1047, these remains must be of early date.[1144] The
gatehouse keep is probably an economical device for combining a citadel
with the defence of the weakest part of the castle.

We must pass on to the keeps of Henry I.[1145] There is only one
in England which authentic history gives to his time, that of
Rochester.[1146] But the chronicler Robert de Torigny[1147] has
fortunately given us a list of the keeps and castles built by Henry in
Normandy, and though many of these are now destroyed, and others in
ruins, a certain number are left, which, taken along with Rochester,
may give us an idea of the type of keep built in Henry I.'s time. The
keeps attributed by Robert to Henry I. are Arques, Gisors, Falaise,
Argentan, Exmes, Domfront, Ambrières, Vire, Waure, Vernon, Evreux,
Alençon, St Jean, and Coutances. How many of these survive we cannot
positively say;[1148] we can only speak of those we have seen (Falaise,
Domfront, and Gisors),[1149] and of Arques, described by M. Deville
in his _Histoire du Chateau d'Arques_, by M. Viollet le Duc in his
treatise on Donjons,[1150] and by Mr G. T. Clark.[1151]

Speaking under correction, as a prolonged study of the keeps in
Normandy was impossible to the writer, we should say that there is no
very striking difference to be observed between the keeps of Henry I.
and those built by his father. The development of the forebuilding
seems to be the most important change, if indeed we are justified
in assuming that the 11th-century keeps never had it; its remains
can be seen at Arques, Falaise,[1152] Domfront, and Rochester. At
Arques and Falaise the doorway is on the second floor, which is an
innovation, a new attempt to solve the difficulty of defending the
entrance. The first floor at Arques could only be entered by a trap
from the second floor; at Falaise there is a stone stair from one to
the other. Rochester is entered from the first floor. The basement
storeys of Arques, Falaise, and Domfront are quite unlit; at the Tower
the basement has had a number of loopholes, and the angular heads of
those which remain suggest that they are at least copied from original
lights. The main floors in Henry I.'s keeps are always of wood, but
this was not because vaulting was then unknown, because the crypt,
sub-crypt, and chapel of the Tower are vaulted, not to speak of many
early churches.[1153] The four keeps mentioned have all three storeys,
thus not exceeding Colchester in height;[1154] the Tower has now four
storeys, but a good authority has remarked that the fourth storey has
not improbably been made by dividing the third.

No marked advance is observable in the masonry of these keeps. Arques
is built of _petit appareil_; Falaise of small stones in herring-bone
work; Domfront of very small stones rudely coursed; Rochester of
Kentish rag mixed with flint rubble. Both Falaise and Domfront have
plinths of superior masonry, but there is always the possibility
that these plinths are later additions. The voussoirs of the arches
at Falaise, Domfront, and Rochester are larger than the rag or tile
voussoirs which are used at Colchester, the Tower, and Langeais. At
Rochester and Arques provision is made for carrying the water-supply
from the well in the basement to the upper floors, a provision of
which there is no trace in the older keeps.[1155]

As Robert de Monte says that Henry I. built many castles in England as
well as in Normandy, we naturally ask what other English keeps besides
Rochester may be assigned to him. It appears to the writer that Corfe
and Norwich keeps may very likely be his. Both were royal castles in
his time, and both were originally wooden castles on mottes.[1156]
Both these castles have forebuildings, and neither of them have
floors supported on vaults.[1157] Corfe has very superior masonry, of
larger stones than those used in the keeps known to be Henry I.'s,
but wide-jointed. At Norwich only a very small piece of the original
ashlar is left. Corfe is extremely severe in all its details, but quite
corresponds to work of Henry I.'s reign.[1158] Norwich has a great deal
of decoration, more advanced in style than that to be seen at Falaise,
but still consistent with the first half of the 12th century. Neither
keep has the least sign of Transition Norman, such as we seldom fail to
find in the keeps of Henry II. Moreover, neither of them figure in the
_Pipe Rolls_ of Henry II., except for repairs; and as Stephen in his
harassed reign can hardly have had any money for building stone keeps,
we may with some confidence ascribe these two keeps to Henry I.

A few words should be given to the castle of Gisors, which contains
in itself an epitome of castle history. The first castle, built by
William Rufus in 1096, was undoubtedly a wooden castle on a motte,
with a stockaded bailey below it; certain portions of the present
bailey walls rest on earthen banks, which probably belonged to the
original castle, and show what a much smaller affair it was than the
present one. Henry I., Robert de Monte tells us, strengthened this
castle with a keep. Probably this was the shell wall which now crowns
the motte; the smallness of the masonry (stones about 5 inches high,
rudely dressed and coursed) and the slight projection of the buttresses
(9 inches) agree with much of the work of his time. There would be a
wooden tower inside.[1159] The chemise or shell wall is pierced by
loopholes, a very unusual arrangement; they are round arched, and of
very rude voussoirs.[1160] Inside this shell there is a decagonal
tower, called the Tower of Thomas à Becket, which is almost certainly
the work of Henry II.,[1161] as its name would indicate; the chapel of
St Thomas is close to it. A stair turret of the 15th century has been
added to this keep; its original entrance was, as usual, a door on the
first floor, but a basement entrance was built afterwards, probably
in the 13th century. Philip Augustus, after he had taken this castle
from John, added to it one of the round keeps which had then become the
fashion, and subsequent enlargements of the bailey converted it into a
"concentric" castle, of which the motte now forms the centre.

There is one keep which is known to be of the reign of Stephen, though
not built by him, that of Carlisle, built by David, King of Scotland,
in 1136,[1162] a time when he thought his hold on the four northern
counties of England was secure, little reckoning on the true character
of his great-nephew, Henry, son of Matilda. There is no advance to be
seen in this keep on those of Henry I., except that the walls are faced
with ashlar. The vaulting of the basement is pronounced by Mr Clark to
be very evidently a late insertion.[1163]

With the reign of Henry II. a new era opens as regards the documentary
history of our ancient castles, because the _Pipe Rolls_ of that king's
reign have most fortunately been preserved.[1164] These contain the
sheriff's accounts for money spent on the building or repair of the
king's castles, and are simply invaluable for the history of castle
architecture. The following is a list of the keeps which the _Pipe
Rolls_ show to have been built or finished by Henry II.:--

 Scarborough, built between 1157 and 1174.  Tower
 Windsor,       "      "    1161  "  1177.  Shell wall
 Orford,        "      "    1165  "  1172.  Tower
 Bridgenorth,   "      "    1166  "  1173.  Tower
 Newcastle,     "      "    1167  "  1177.  Tower
 Bowes,         "      "    1171  "  1187.  Tower
 Richmond,      "      "    1171  "  1174.  Tower
 Chilham,       "      "    1171  "  1173.  Tower
 Peak,          "      "    1172  "  1176.  Tower
 Canterbury,    "      "    1172  "  1174.  Tower
 Arundel,       "      "    1176  "  1182.  Shell wall
 Tickhill,      "      "    1178  "  1181.  Tower

The dates given here must be taken as only roughly accurate, as owing
to the meagreness of the entries in the _Pipe Rolls_, it is not always
certain whether the expenses were for the great tower or for other
buildings. The list by no means includes all the work which Henry II.
did on his English castles, for he was a great builder; but a good
deal of his work seems to have been the substitution of stone walls
with mural towers, for wooden stockades, and our list comprises all
the cases in which it is clear that the keep was the work of this
king.[1165] We confine our attention to the keeps, because though mural
towers of stone began to be added to the walls of baileys during Henry
II.'s reign (a detail which must have greatly altered the general
appearance of castles), it is certain that the keep was still the most
important part, and the residence of the king or noble whenever he
visited the castle.

Seven out of the ten tower keeps are built on precisely the same plan
as those of Henry I. The chief advance is in the masonry. All the tower
keeps of Henry II., except Dover, Chilham, and Canterbury, are or have
been cased with good ashlar, of stones somewhat larger in size than
those used by Henry I. The same may be said of the shell walls (namely,
Windsor and Arundel); it is interesting to note that Henry II. still
used this elementary form of citadel, which consisted merely of a wall
round the top of a motte, with wooden buildings inside.[1166] In three
cases out of the ten tower keeps, Newcastle, Bowes, and Richmond,
the basement storey is vaulted, which does not occur in the older
keeps.[1167] Yet such important castles as Scarborough, Dover, and
Canterbury are without this provision against fire. None of these keeps
appear to have more than three storeys above the basement.[1168] None
of the entrances to the keeps (except Tickhill) have any portcullis
grooves,[1169] nor any special contrivances for defence, except at
Canterbury, where the entrance (on the first floor) takes two turns at
right angles before reaching the hall to which it leads.[1170] There
are nearly always in the keeps of Henry II. some signs of Transition
Norman in the details, such as the nook shafts at the angles of the
towers of Scarborough and the Peak, certain arches at Canterbury, the
Transition capitals used at Newcastle, and the filleted string round
the outside of Bowes.

But we have yet to speak of three keeps of Henry II.'s reign which
are on a different plan to all the others, and which point to coming
changes--Chilham, Orford, and Tickhill.[1171] Chilham is an octagonal
tower of three storeys, with a square annexe on one side, which
appears to be original. Orford is polygonal outside, round inside.
Orford indeed is one of the most extraordinary keeps to be seen
anywhere, and we must regard it as an experiment, and an experiment
which appears never to have been repeated.[1172] Instead of the usual
Norman buttresses, this polygonal keep has three buttress towers,
placed between every four of the outer faces, 22 feet wide, and 12
feet in projection.[1173] Tickhill, however, the last keep he built,
is decagonal. The object of the polygonal tower was to deflect the
missiles thrown from siege engines, and the round tower was evidently
considered an improvement on the polygonal for this purpose, as it
subsequently supplanted the polygonal type. It is therefore rather
remarkable that Henry II. built both these keeps in the second decade
of his reign, and afterwards went on building square keeps like his
predecessors. We have seen, however, that he built at least one
polygonal tower in Normandy, that of Gisors. We must bear in mind that
the Norman and Angevine frontier was the theatre of the continuous
struggle of Henry II. with the French kings, Louis VII. and Philip
Augustus, and that it is here that we must expect the greatest
developments in military architecture.

Speaking generally, we may say that just as there was comparatively
little change in armour during the 12th century until the end of Henry
II.'s reign, so there was comparatively little change in military
architecture during the same period. But great changes took place
towards the end of the 12th century. One of these changes was a great
improvement in missile engines; the trébuchet was one of the most
important of these. It could throw much heavier stones than the largest
catapult, and could take a more accurate aim.[1174] These new engines
were useful for defence as well as attack, and this affected the
architecture of castles, because flat roofs covered with lead, on which
machines could be placed, were now substituted for the former sloping
roofs.[1175] There are several payments for lead for roofing castles
in the _Pipe Rolls_ of Henry II., the earliest being in 1166. In the
reigns of John and Henry III. the mention of lead for roofing becomes
much more frequent.[1176]

Hitherto, in the defence of keeps, reliance had mainly been placed upon
their passive strength, though not so entirely as has been commonly
assumed, since it was always the practice to shoot with arrows from the
battlements round the roof of the tower. But not only was the fighting
strength of the keep increased by the trébuchet, but the introduction
of the crossbow gave it a defensive arm of the greatest importance.
The crossbow had been known to the Romans, and was used in the early
part of the 12th century, but it was forbidden by the second Lateran
Council in 1139 as a weapon hateful to God.[1177] This prohibition
seems actually to have been effective, as William the Breton says
expressly that the crossbow was unknown to the French before the wars
of Richard I. and Philip Augustus.[1178] Richard learned the use of it
in the third crusade.[1179] But to use the crossbow in the defence of
buildings it was necessary to construct special loopholes for shooting,
splayed downwards externally, so that it was possible to aim from them.
Up till this time the loopholes of castles had been purely for light
and not for shooting; anyone may see that it is impossible to take aim
through an immensely thick wall unless there is a downward splay to
increase the field of vision. William the Breton tells us that Richard
built windows for crossbows to his towers, and this is the first
mention we have of them.[1180]

From this time defensive loopholes become common in castles, and take
various fanciful forms, as well as the commoner ones of the circle,
square, or triangle at the base of the loop. The _cross loophole_,
which does not appear till the latter quarter of the 13th century, is
explained by Viollet le Duc as an ingenious way of allowing three or
four archers to fire in a volley.[1181] But up to the present time very
little study has been given to this subject, and we must be content to
leave the question for future observation to settle.[1182]

The crossbowmen not only required splayed loopholes, but also niches,
large enough to accommodate at least three men, so that a continuous
discharge of darts (quarrells) might be kept up. Any defensive loop
which really means work will have a niche like this behind it. These
niches had the defect of seriously weakening the wall.

Another innovation introduced by Richard I. was that of stone
machicolations, or _hurdicia_.[1183] Whether wooden galleries round the
tops of walls, with holes for dropping down stones, boiling-water, or
pitch on the heads of the besiegers had not been used from the earliest
times, is regarded by Köhler as extremely doubtful.[1184] They were
certainly used by the Romans, and may even be seen clearly figured on
the Assyrian monuments. In the Bayeux Tapestry, the picture of Bayeux
Castle shows the stockade on top of the motte crested with something
extremely like hurdicia. Yet the writer has found no authentic
mention of them before the end of the 12th century.[1185] The stone
machicolations built by Richard round his keep of Chateau Gaillard are
of an unusual type, which was only rarely imitated.[1186] But from
this time wooden hurdicia became universal, to judge from the numerous
orders for timber for _hoarding_ castles and town walls in the _Close
Rolls_ of the first half of the 13th century. Towards the middle of the
13th century stone brackets for the support of wooden hurdicia began
to be used; they may still be seen in the great keep of Coucy, which
was begun in 1230. But machicolations entirely of stone, supported on
double or triple rows of brackets, do not become common till the 14th
century.[1187]

The greatest architectural change witnessed at the end of the 12th
century was the victory of the round keep over the square. Round towers
were built by the Romans as mural towers, but the universal type of
mediæval keep appears to have been the square or oblong, until towards
the end of the 12th century.[1188] The polygonal keep was probably a
transitional form; we have seen that Henry II.'s polygonal keep at
Orford was begun as early as 1165. Many experiments seem to have been
made at the end of the 12th century, such as the addition of a stone
prow to the weakest side of a keep, to enable it better to resist
showers of missiles. Richard I.'s keep at Chateau Gaillard is a round
keep with a solid prow of this kind. Five-sided keeps are said to be
not uncommon on the left bank of the Rhine and in Nassau; this type
was simply the addition of a prow to a square keep. The only English
instance known to the writer is that of Mitford, Northumberland, but
this is merely a five-sided keep, the prow is not solid, as at Chateau
Gaillard. The castle of Étampes, whose plan is a quatrefoil, is
assigned by French archæologists to this period of experiment.[1189]
But the round keep was eventually the type preferred. Philip II.
thought it necessary to add a round keep to the castle of Gisors, after
he had taken it from John, and he adopted the round keep for all his
new castles, of which the Louvre was one.[1190]

Along with the round keep, ground entrances became common.[1191]
Viollet le Duc states that when the French soldiers broke into the
inner ward at Chateau Gaillard the defenders had no time to escape
into the keep by the narrow stair which led to the first floor, and
consequently this proud tower was surrendered without a blow; and
that this event so impressed on Philip's mind the danger of difficult
entrances that he abandoned the old fashion. This may be true, but it
is a pure guess of Le Duc's, as there is nothing whatever to justify
it in William the Breton's circumstantial narrative. It is, however,
certain that Philip adopted the ground entrance to all his keeps.
In England we find ground entrances to many round keeps of the 13th
century, as at Pembroke; but the older fashion was sometimes retained;
Conisburgh, one of the finest keeps in England, has its entrance on the
first floor.[1192]

After the introduction of the trébuchet, we might expect that the
walls of keeps would be made very much thicker, and such seems to
have been the case in France,[1193] but we do not find that it was
the rule in England.[1194] The lower storeys were now generally
instead of occasionally vaulted. In the course of the 13th century it
became common to vault all the storeys. But in spite of the military
advantages of the round keep, in its avoidance of angles favourable to
the battering-ram, and its deflection of missiles, the square keep
continued to be built in various parts of both France and England till
quite late in the Middle Ages.[1195] On the Scottish border, square
towers of the ancient type, with quite Norman decorations, were built
as late as the 15th century.[1196] The advantage of the square tower
was that it was more roomy inside, and was therefore preferred when the
tower was intended for habitation.

We come now to the greatest of all the changes introduced in the
13th century: the keepless castle, in which the keep is done away
with altogether, and the castle consists of a square or oblong court
surrounded by a strong wall with massive towers at the angles, and in
large castles, in the curtain also.[1197] Usually this inner quadrangle
is encircled with an outer quadrangle of walls and towers, so that this
type of castle is frequently called the _concentric_. But the castles
of the keepless kind are not invariably concentric; those built by
Edward I. at Conway, Carnarvon, and Flint are not so.[1198] Instead
of a dark and comfortless keep, the royal or noble owner is provided
in this type of castle with a palatial house. In England this house
is frequently attached to the gateway, forming what we may call a
gatehouse palace; good examples may be seen at Beaumaris, Harlech, and
Tonbridge.[1199] The gateway itself is always defended by a pair of
massive towers.

Edward I. is generally credited with the introduction of this type
of castle into England, but until the _Pipe Rolls_ of Henry III.'s
reign have been carefully examined, we cannot be certain that it was
not introduced earlier. It was certainly known in Germany fifty years
before Edward's accession to the throne, and in France as early as
1231.[1200]

It is always supposed that this type of castle was introduced by the
Crusaders from Syria. But when did it make its first appearance in
Syria? This is a point which, we venture to think, has not been yet
sufficiently investigated. We do not believe that it can have existed
in Syria at the time of the third crusade, otherwise Richard I., who is
universally acknowledged to have been a first-class military architect,
would have brought the idea home with him.[1201] Yet his favourite
castle of Chateau Gaillard, built in accordance with the latest
military science, is in the main a castle of the keep-and-bailey type,
and has even a reminiscence of the motte, in the scarped rock on which
the keep and inner ward are placed.

The new type of keepless castle never entirely displaced the old
keep-and-bailey type. We have already seen that keeps of the old sort
continued to be built till the end of the Middle Ages. Hawarden Castle
has a good example of a 14th-century round keep; Warkworth a most
remarkable specimen of the 15th, the plan being a square tower with
polygonal turrets set on each face.[1202] In France and Germany also
the old type appears to have persisted.[1203]

We have already trespassed beyond the limits of our subject; but as we
offer this chapter more as a programme of work than as a categorical
outline, we trust it may not be without use to the student who may feel
disposed to take up this much-neglected subject.

A few words must yet be said about the state of the law relating to
castles. Nothing explicit has come down to us on this subject from the
11th century in England, but it is clear that the feudal system which
William introduced, and which required that all lands should revert to
the king on the death of the holder, forbade the building of any castle
without the king's license, and, further, allowed only a life tenure
in each case. The Council of Lillebonne in 1080 had laid it down in
express terms that no one should build a castle in Normandy without the
permission of the duke;[1204] and William, after his great victory
over his revolted barons, had enforced the right of garrisoning their
castles. He was not able to do this in England, while he must have
desired to check the building of private castles as far as possible. On
the other hand, he had to face the dilemma that no Norman land-holder
would be safe in his usurped estates without the shelter of a castle.
In this situation we have the elements of the civil strife which burst
forth in Stephen's reign, and which was ended by what we may call the
anti-castle policy of Henry II.[1205]

The rights secured by this able king were often recklessly sold by his
successors, but in the reign of Henry III. it was evidently illegal
even to fortify an ordinary house with a ditch and stockade without
royal permission.[1206]

       *       *       *       *       *

Feudalism was an inevitable phase in the evolution of the Western
nations, and it ought neither to be idealised nor execrated. After
the break-up of the tribal system the nations of Europe sought refuge
in the forms of imperialism which were devised by Charlemagne, and
even the small and distant island of England strove to move in the
same direction. But the times were not ripe for centralisation on
so great a scale, and when the system of the Carlovingian Empire
gave way under the inrush of Northmen and Huns, European society
would have fallen into ruin had it not been for the institutions of
feudalism. These offered, in place of the old blood bond of the tribe,
a social compact which, though itself artificial, was so admirably
adapted to the general need that it was speedily adopted by all the
progressive nations of Europe. The great merit of feudalism was that it
replaced the collective responsibility of the tribe by the individual
responsibility of the man to his lord, and of the lord to his man. In
an age when the decay of mutual trust was the worst evil of society it
laid stress on individual loyalty, and insisted that personal honour
should consist in the fulfilment of obligations. Being a system so
wholly personal, its usefulness depended largely on the nature of the
person in power, and it was therefore liable to great abuses.

But it is probable that feudalism worked better on the whole in
England than in any other part of Western Europe. The worst evils of
French feudalism never appeared in this country, except during the
short and disastrous reign of Stephen. The strong kings of the Norman
and Plantagenet Houses held in check the turbulence of the barons;
and private war was never allowed to become here, as it was on the
Continent, a standing evil. To follow out this subject would lead us
beyond the limits of this book, but it is interesting to remember that
not only the picturesque ruins of our castles, but also the neglected
green hillocks of which we have treated in this work, while they point
to the skilful machinery by which the Norman Conquest was riveted on
the land, bear witness also to something still more important. They
tell of a period of discipline and education through which the English
people passed, when in spite of much oppression and sometimes even
cruelty, seeds of many noble and useful things were sown, from which
succeeding generations have garnered the enduring fruit.




APPENDICES




APPENDIX A

PRIMITIVE FOLK-MOOTS


The popular meetings of the Anglo-Saxons, those of the hundred and the
shire, were held in the open air. Since many of those who attended them
had to travel far, some sign was necessary to mark out the place of
meeting, and some striking feature, such as a hillock, or a particular
tree, or an ancient barrow, was chosen. Thus we have the Shire Oak,
near Leeds, which gives its name to the wapentake of Skyrack; and in
a charter of Edgar we find the _mot-beorh_ mentioned, and translated
_Congressionis Collem_ = the meeting barrow. (_M. A._, ii., 324.)
It does not appear that a hillock was an essential feature of these
meeting-places, though this is popularly supposed to be the case,
because the "Thing-wall" in Iceland and the "Tynwald" in the Isle of
Man have hillocks from which laws were proclaimed. The Thingwall, or
field of meeting in Iceland had a natural rock just above it, isolated
by a stream, and though proclamations were made from this rock,
deliberations took place on the level. (Gomme's _Primitive Folk-Moots_,
31.)

The Tynwald Hill, in the Isle of Man, which is also still used for
the proclamation of new laws, was probably an ancient barrow, as
there are other barrows in the immediate neighbourhood. (Kermode and
Herdman, _Illustrated Notes on Manx Antiquities_, pp. 23 and 61.) At
Thingwall, near Liverpool, and Thingwall in Wirral, both probably Norse
settlements, there is no hillock.

In Scotland, the use of a former motte as a meeting-place for the
baronial court appears to have been much more common than in England.
Mr George Neilson's explanation of this fact is referred to in Chapter
X., p. 307.




APPENDIX B

WATLING STREET AND THE DANELAGH


It has been pointed out by Schmid (_Gesetze der Angelsachsen_,
xxxviii.) that the document called _Alfred and Guthrum's Peace_ cannot
belong to the year of Guthrum's baptism at Wedmore; and Mr J. R. Green
(_Conquest of England_, p. 151) goes further, and doubts whether the
boundaries laid down in this deed refer to anything except to the East
Anglian kingdom of Guthrum. But Mr Green gives no adequate reason for
rejecting the generally accepted conclusion that the Watling Street was
the boundary between English and Danish Mercia, which is borne out by
the following facts: (1) the Danish confederacy of the five boroughs,
Lincoln, Stamford, Leicester, Nottingham, and Derby, pretty well covers
the part of Mercia north of Watling Street, especially when Chester
is added, as it sometimes is, to the list; (2) the division into
wapentakes instead of hundreds, now believed to be of Danish origin,
is found in Lincolnshire, Notts, Derbyshire, Rutland, Leicestershire,
and Northamptonshire. Staffordshire, it is true, is not divided into
wapentakes, but it was apparently won by conquest when Ethelfleda
fortified the town. Chester was occupied by her husband in 908. Watling
Street furnishes such a well-defined line that it was natural to fix
upon it as a frontier.




APPENDIX C

THE MILITARY ORIGIN OF ALFRED'S BOROUGHS


Keutgen (_Untersuchungen über den Ursprung der Deutschen
Stadtverfassung_, 1895) appears to have been the first to notice the
military origin of the Old Saxon boroughs; and Professor Maitland saw
the applicability of the theory to the boroughs of Alfred and Edward
the Elder. (_Domesday Book and Beyond._) The _Anglo-Saxon Chronicle_,
in 894, speaks of "the men whose duty it was to defend the towns"; this
proves that Alfred had made some special arrangement for the defence of
the towns; and this arrangement must have been something quite apart
from the ordinary service of the _fyrd_ or militia, which was only
due for a short time. It must have been something permanent, with an
adequate economic basis, such as we have in Henry the Fowler's plan.




APPENDIX D

THE WORDS "CASTRUM" AND "CASTELLUM"


If we take the chroniclers of the reign of Charlemagne and his
successors in the 9th century, we find the word _castrum_ constantly
used for places such as Avignon, Dijon, Macon, Rheims, Chalons,
Cologne, Andernach, Bonn, Coblenz, etc., all of which are known to
have been Roman _castra_, when there can be no doubt that the _city_
is meant. Take, for instance, the _Annales Mettenses_ (Pertz, i.,
326), 737: Karl Martel hears that the Saracens have taken "castrum
munitissimum Avinionem" (Avignon); he marches against them, and
"_predictam urbem_ obsidione circumdat." But these cities are not only
called _castra_, they are also called _castella_. Thus the chronicle
ascribed to Hincmar calls Macon both _castrum_ and _castellum_ in
the same breath. (_Migne_, 125, 1298.) The fortifications built by
Charlemagne against the Saxons are called _castra_, _castella_, and
_civitates_. (_Chron. Moissiacense_, Pertz, i., 308. _Ann. Einhardi_,
ibid., 196, 204.) The camps of the Northmen, which as we have seen,
were of great size, are also called not only _castra_, but _civitates_,
_castella_, _munitiones_, _oppida_. (_Annales Fuldenses_, Pertz, i.,
397.) The camp built by Charles the Bald at Pistes in 868 is called a
_castellum_, though it was evidently an enclosure of great size, as he
measured out quarters in it for his nobles, and formed an elaborate
scheme for its maintenance. (Hincmar, _Migne_, 125, 1242, 1244.) Coming
to the 10th century, the following passage from Flodoard will show the
vagueness of the words in common use for fortifications: "Heribertus
Ansellum Bosonis subditum, qui prædictum custodiebat _castrum_ (Vitry),
cum ipso _castello_ recipit, et Codiacum S. Remigii _municipium_ illi
cum alia terra concedit. Nec longum, Bosonis fideles _oppidanorum_
proditione Victoriacum (Vitry) recipiunt, et Mosonum fraude pervadunt.
At Heribertus, a quibusdam Mosomensibus evocatus, supervenit
insperatus, et entrans _oppidum_, porta latenter a _civibus_ aperta,
milites Bosonis, qui ad custodiam loci residebant, ibidem omnes capit."
(_Migne_, 135, 297.) Here it is clear that _castrum_, _castellum_,
_municipium_, and _oppidum_ all mean the same thing, and the one word
_civibus_ betrays that it is a _city_ which is meant. Undoubtedly the
chronicler thinks it elegant to change his words as often as he can.
_Munitio_ is another word frequently used; in classical Latin it means
a bulwark, a wall or bank; in the chroniclers of the 10th century it
is used indifferently for a town or castle, though certain passages,
such as "subversis multarum munitionibus urbium" (Flodoard, i., vi.),
show that the right sense is not far from the mind of the writer. The
numerous passages in which we are told of monasteries being enclosed
with walls and converted into _castella_, show that the _enclosure_
is the chief idea which the chroniclers associate with this word. The
citations made above are not exceptional, but typical, and could be
paralleled by countless others.

Since the above was written, I have read Keutgen's _Untersuchungen über
den Ursprung der Deutschen Stadtverfassung_. He remarks that the Latin
words for a town (in the 10th and 11th century writers) are _urbs_,
_castellum_, _civitas_, sometimes _arx_; for a village, _villa_,
_oppidum_, _vicus_. This absolutely agrees with what I have observed in
these writers, except that I have certainly found _oppidum_ used for a
town, as in the passage from Flodoard cited above.




APPENDIX E

THE BURGHAL HIDAGE


The _Burghal Hidage_ has been printed by Birch, _Cartularium_, iii.,
671. The manuscript is very corrupt, and several of the places
cannot be identified. Those which can be identified are: Hastings,
Lewes, Burpham (near Arundel), Chichester, Porchester, Southampton,
Winchester, Wilton, Tisbury, Shaftesbury, Twineham, Wareham, Bridport,
Exeter, Halwell, Lidford, Pilton, Barnstaple, Watchet, Axbridge,
Lyng (near Athelney), Langport, Bath, Malmesbury, Cricklade, Oxford,
Wallingford, Buckingham, Eashing (near Guildford), and Southwark. The
list thus seems to give an outline of Alfred's kingdom as it was at
his death, or at the beginning of the reign of his son. Dr Liebermann
refers it to the latter date. (_Leges Anglorum_, 9.)




APPENDIX F

THELWALL


A writer in the _Manchester Guardian_ a few years ago suggested a
new solution of the name Thelwall. He believes that the Thelwall
raised by Edward was a boundary wall of timber, stretching from
Thelwall to Runcorn. The Mersey, he argues, above Thelwall formerly
broadened out into a series of swamps which would effectually defend
the frontier towards the east. But westward from Thelwall there were
no such obstacles, and it is assumed that Edward made a timber wall
from Thelwall to Ethelfleda's fortress at Runcorn. Some support to
this hypothesis is given in the names of places between Thelwall and
Runcorn: Stockton, Walton (twice), Stockham, Walford, Wallmore, and
Wall-hes. Further, when the bed of the Mersey was delved for the Ship
Canal, discovery was made of "a remarkable series of submerged piles,
9 feet long, arranged in two parallel ranks which were 30 feet apart.
The intervals between the piles varied, but seem to have averaged 5 to
6 feet. Between the ranks were diagonal rows of upright stakes, each
stake about 5 feet long, extending from either rank chevron-wise to the
middle and there overlapping, so that the ground-plan of them makes
a kind of herring-bone pattern. By this plan, anyone passing through
would have to make a zigzag course. In some places sticks and sedges
were found interwoven horizontally with the stakes, a condition of
things which probably obtained throughout the whole series. The tops
of the tallest piles were 10 feet below the present surface of the
ground, which fact goes far toward precluding the possibility that this
elaborate work may have been a fish-weir. The disposition of the stakes
points to a military origin. So arranged, the advantage they offered
to defending forces was enormous." I think it worth while to reproduce
this account, especially because of the place-names, but those who are
learned in the construction of fish-weirs may perhaps think that the
description will apply to a work of that kind.




APPENDIX G

THE WORD "BRETASCHE"


This word, which also appears as bretagium, britagium, or bristega,
evidently means a tower, as is clear from the following passages: Order
from King John to erect a _mota et bretagium_ at Roscrea, in Ireland
(Sweetman's _Calendar_, i., 412); Order by Henry III. to the dwellers
in the Valley of Montgomery "quod sine dilatione motas suas bonis
bretaschiis firmari faciant" (_Close Rolls_, ii., 42); Order that the
timber and bretasche of Nafferton Castle be carried to Newcastle, and
the bretasche to be placed at the gate of the drawbridge _in place of
the little tower_ which fell through defect in its foundations (_Close
Rolls_, i., 549b).

The word is also expressly defined by William the Breton as a wooden
castle: "Circuibat castrum ex omni parte, et fabricavit brestachias
duplices per septem loca, _castella videlicet lignea_ munitissima."
(Bouquet, xvii., 78.)

See also Wright, "Illustrations of Domestic Architecture," _Arch.
Journ._, i., 212 and 301. In these papers it is clear that "breteske"
means a tower, as there are several pictures of it. At a later period
it seems to have been used for a wooden balcony made for the purpose of
shooting, in the same sense as the word "hurdicium"; but I have not met
with any instance of this before the 14th century.




APPENDIX H

THE WORDS "HURDICIUM" AND "HORDIARI"


These words refer to the wooden galleries carried round the tops of
walls, to enable the defenders to throw down big stones or other
missiles on those who were attempting to attack the foot of the walls.
"Hurdicia quæ muros tutos reddebant." (_Philippidos_, vii., 201;
Bouquet, xvii.) The word "alures" is sometimes used in the same sense.
See a mandamus of Henry III., cited by Turner, _History of Domestic
Architecture_, i., 198: "To make on the same tower [of London] on the
south side, at the top, deep alures of good and strong timber, entirely
and well covered with lead, through which people can look even to the
foot of the tower, and better defend it, if need may be." The alures of
the castle of Norwich are spoken of as early as 1187, but this mention,
and one of the alures round the castle of Winchester in 1193, are the
only ones I find in the 12th century in England.




APPENDIX I

"HERICIO, ERICIO, HERITO, HERISSON"


This is derived from the French word _hérisson_, a hedgehog, and should
mean something bristling, perhaps with thorns or osiers. Several
passages show that it was a defence on the counterscarp of the ditch,
and it may sometimes have been a hedge. Cohausen, _Befestigungen
der Vorzeit_, shows that hedges were frequently used in early
fortifications (pp. 8-13). The following passages seem to show clearly
that it was on the counterscarp of the ditch: "[Montreuil] il a bien
clos, esforce e ferme de pel e _hericon_." (Wace, 107.) "Reparato
exterioris Ardensis munitionis valli fossato et amplificato, et sepibus
et ericiis consepto et constipato." (Lambert of Ardres, 623, _circa_
1117.) The French poem of Jordan Fantosme, describing the siege of
Wark by the Scots in 1174, says the Scots attacked and carried the
_hericon_, and got into the ditch, but they could not take the bayle,
_i.e._, they could not get over the palicium.




APPENDIX K

THE CASTLE OF YALE


In the year 1693, the antiquary Edward Llwyd was sitting on the motte
of Tomen y Rhoddwy engaged in making a very bad plan of the castle
[published in _Arch. Camb._, N.S., ii., 57]. His guide told him that he
had heard his grandfather say that two earls used to live there. Llwyd
called the guide an ignorant fellow. Modern traditions are generally
the work of some antiquary who has succeeded in planting his theories
locally; but here we have a tradition of much earlier date than the
time when antiquaries began to sow tares, and such traditions have
usually a shred of truth in them. Is it possible that this castle of
Tomen y Rhoddwy and the neighbouring one of Llanarmon were built by
the earls of Chester and Shrewsbury, who certainly went on expeditions
together against Wales, and appear to have divided their conquests?
It is to be noted that the township is called _Bodigre yr Yarll_, the
township of the earls.




APPENDIX L

THE CASTLE OF TULLOW OR "COLLACHT," p. 335


This information is kindly supplied by Mr Goddard H. Orpen, who writes
to me: "I visited Tullow lately, and asked myself where would a Norman
erect a mote, and I had no difficulty in answering: on the high ground
near where the Protestant church stands. When I got up there the first
thing that I noticed was that the church stood on a platform of earth
10 to 14 feet higher than the road, and that this platform was held in
position by a strong retaining wall, well battered towards the bottom
on one side. I then found on enquiry that the hill on which it stood
and the place to the N.W. of it was called the 'Castle Hill.' On going
round to the N.W. of the church I found a horseshoe-shaped space,
scarped all round to a height of 6 to 10 feet, and rising to about 16
feet above the adjoining fields. There is no doubt that this was the
site of the castle, and that it was artificially raised. To my mind
there was further little doubt that it represented an earlier mote. In
a field adjoining on the W. I could detect a platform of about 50 to
70 paces, with traces of a fosse round the three outer sides.... This
was certainly the Castellum de Tulach mentioned in the deeds concerning
Raymond le Gros' grant to the Abbey of St Thomas.--_Dublin Reg. St
Thomas_, pp. 111, 113."




APPENDIX M

THE CASTLE OF SLANE


Mr Westropp says that the "great earthworks and fosses" on the Hill of
Slane are mentioned in the "Life of St Patrick" (_Journ. R. S. A. I._,
1904, p. 313). What the _Life_ really says is: "They came to Ferta Fer
Fiecc," which is translated "the graves of Fiacc's men"; and the notes
of Muirchu Maccu-Machtheni add, "which, as fables say, were dug by the
slaves of Feccol Ferchertni, one of the nine Wizards" (_Tripartite
Life_, p. 278). It does not mention any fort, or even a hill, and
though Ferta Fer Fiecc is identified with Slane, there is nothing to
show what part of Slane it was.




APPENDIX N

THE WORD "DONJON"


Professor Skeat and _The New English Dictionary_ derive this word
from the Low Latin, _dominionem_, acc. of _dominio_, lordship. Leland
frequently speaks of the keep as the dungeon, which of course is the
same word. Its modern use for a subterranean prison seems to have
arisen when the keeps were abandoned for more spacious and comfortable
habitations by the noble owners, and were chiefly used as prisons.
The word _dunio_, which, as we have seen, Lambert of Ardres used
for a motte, probably comes from a different root, cognate with the
Anglo-Saxon _dun_, a hill, and used in Flanders for the numerous
sandhills of that coast.




APPENDIX O

THE ARRANGEMENTS IN EARLY KEEPS


We get a glimpse of these in a story given in the "Gesta Ambasiensium
Dominorum," D'Archery, _Spicilegium_, 278. Sulpicius the Treasurer
of the Abbey of St Martin at Tours, an important personage, built a
stone keep at Amboise in 1015 (_Chron. Turonense Magnum_), in place
of the "wooden house" which his brother had held. In the time of Fulk
Rechin (1066-1106), this keep was in the hands of the adherents of
the counts of Blois. Hugh, son of Sulpicius, with two other men, hid
themselves by night in the basement, which was used as a storehouse;
it must therefore have had an entrance from outside. With the help
of ropes, they climbed up a sewer into the bedchamber, which was
above the cellar, and evidently had no stair communicating with the
cellar. Here they found the lady of the house and two maids sleeping,
and a watchman who was also asleep. While one of the men held these
in terror with a drawn sword, the other two climbed up a ladder and
through a trap-door up to the roof of the tower, where they unfurled
the banner of Hugh. Here we see a very simple keep, which has only one
storey above the basement; this may have been divided into two or more
apartments, but it was thought a fitting residence for a lady of rank.
It had no stairs, but all the communications were by trap-doors and
ladders. We may be quite sure that the people of rank of the 11th and
12th centuries were content with much rougher accommodation than Mr
Clark imagined. Even Richard I.'s much admired keep of Chateau Gaillard
appears to have had no communication but ladders between the floors.




APPENDIX P

KEEPS AS RESIDENCES


The description of a keep which we have already given from Lambert of
Ardres (Chap. VI.) is sufficient to prove that even wooden keeps in the
12th century were used as permanent residences, and this is confirmed
by many scattered notices in the various chronicles of France and
England. It was not till late in the 13th century that the desire for
more comfortable rooms led to the building of chambers in the courtyard.




APPENDIX Q

CASTLES BUILT BY HENRY I.


The castles, which according to Robert de Monte, Henry I. built
altogether [_ex integro_] were Drincourt, Chateauneuf-sur-Epte,
Verneuil, Nonancourt, Bonmoulins, Colmemont, Pontorson, St
Denis-en-Lyons, and Vaudreuil. Many of these may have been wooden
castles; Chateauneuf-sur-Epte almost certainly was; it has now a
_round_ donjon on a motte. The "Tour Grise" at Verneuil is certainly
not the work of Henry I., but belongs to the 13th century.




APPENDIX R

THE SO-CALLED SHELL KEEP


We have three accounts of motte-castles from the 12th century: that
of Alexander Neckham, in the treatise _De Utensilibus_; that of
Laurence of Durham, cited in Chapter VII., p. 147; and the well-known
description of the castle of Marchem, also cited in Chapter VI., p.
88. All these three describe the top of the motte as surrounded by a
wall (of course of wood), within which is built a wooden tower. The
account of Marchem says that it was built in the middle of the area.
This supports the conjecture in the text. Mr H. E. Malden has shown
(_Surrey Archæolog. Collections_, xvi., 28) that the keep of Guildford
is of later date than the stone wall round the top of the motte. Remove
this tower, and there would be what is commonly called a shell keep. It
would appear, therefore, that it was a common practice to change the
bank or stockade round the top of the motte into a stone wall (no doubt
as a defence against fire), leaving the keep inside still of wood. Four
of the pictures from the Bayeux Tapestry (see Frontispiece) all give
the idea of a wooden tower inside a stockade on a motte.




APPENDIX S

PROFESSOR LLOYD'S "HISTORY OF WALES"


I regret that this valuable work did not appear until too late for me
to make use of it in my chapter on Welsh Castles. It is worth while to
note the following points in which Professor Lloyd's conclusions differ
from or confirm those which I have been led to adopt.

Aberystwyth and Aberrheiddiol.--"After the destruction of the last
Aberystwyth Castle of the older situation in 1143, the chief
stronghold of the district was moved to the mouth of the Rheiddiol,
a position which it ever afterwards retained, though people still
insisted on calling it Aberystwyth" (514). "The original castle of
Aberystwyth crowned the slight eminence at the back of the farm of Tan
y Castell, which lies in the Ystwyth valley 1-1/2 miles S. of the town.
There is the further evidence of the name, and the earthworks still
visible on the summit" (426, _note_).

Carreghova.--I ought perhaps to have included this castle in my list,
though on the actual map its site is within the English border; but as
there are absolutely no remains of it [D. H. M.] it does not affect the
question I am discussing.

Cardigan and Cilgerran.--"Dingeraint cannot be Cilgerran, because
Cilgerran is derived from _Cerran_, with the feminine inflection, not
from _Geraint_; nor is Cilgerran 'close to the fall of the Teifi into
the sea,' as the chronicler says Dingeraint was. The castle built by
Earl Roger was probably Cardigan" (401). Professor Lloyd afterwards
identifies Cilgerran with the castle of Emlyn (661). This seems to me
questionable, as the "New Castle of Emlyn," first mentioned in Edward
I.'s reign, presupposes an older castle, and as I have stated, a mound
answering to the older castle still exists not far from the stone
castle.

Carmarthen.--Professor Lloyd thinks this castle stood at the present
farm of Rhyd y Gors, about a mile below the town; but I see no reason
to alter the conclusion to which I was led by Mr Floyd's paper, that
the Rhyd y Gors of the castle was a ford at Carmarthen itself. The fact
that Henry I. founded a cell to Battle Abbey at Carmarthen (431) seems
to me an additional piece of evidence that the castle was there; castle
and abbey nearly always went together.

Dinweiler.--Professor Lloyd assumes Dinweiler to be the same as the
castle in Mabudryd built by Earl Gilbert, and to be situated at or
near Pencader (501). It should be noted, however, that Dinweiler reads
Dinefor in MS. B. of the _Brut_, in 1158. I am in error in supposing St
Clair to be the castle of Mabudryd (following a writer in _Archæologia
Cambrensis_), as St Clair is not in that commote. Professor Lloyd's map
of the _cantrefs_ and _commotes_ differs widely from that of previous
writers.

Llangadoc.--"Luchewein" should not be identified with this castle;
Professor Lloyd thinks it may refer to a castle at Llwch Owain, a lake
in the parish of Llanarthney, where there is an entrenchment known as
Castell y Garreg.

Maud's Castle.--Camden identified "Matildis castrum" with Colewent or
Colwyn, but Professor Lloyd is of opinion that "a careful collation of
the English and Welsh authorities for the events of the years 1198 and
1231 will make it clear that Payne's Castle and Maud's Castle are the
same." This of course does not affect what is said about Colwyn Castle
in the text.

Montgomery.--Professor Lloyd deems that the emphasis laid (especially
in the _Charter Rolls_, i., 101) on the fact that the building of Henry
III.'s reign was New Montgomery, leaves no doubt that the former town
and castle stood elsewhere, probably at Hên Domen. This, if true, would
greatly strengthen my case, as Hên Domen is an admirable motte and
bailey.




SCHEDULE OF ENGLISH CASTLES KNOWN TO DATE FROM THE ELEVENTH
CENTURY[1207]


IN TOWNS

      Name of                      Whole Area of Enceinte
 No.  Castle.     Type.[1208]        or Bailey.               Value.

  1. Arundel      M. and B., O.    Whole area 4-1/2 acres    Risen.
  2. Bamborough   K. and B.        Whole area 4-3/4 acres     ...
  3. Barnstaple   M. and B.        Bailey 1-1/3 acres        Not given
                                                              T. R. E.
  4. Bristol      M. and B., O.    Whole area nearly 4 acres Not given
                                                              T. R. E.
  5. Buckingham   M. and B.                   ?              Risen.
  6. Caerleon     M. and B., O.    Bailey 4-3/4 acres        Risen.
  7. Cambridge    M. and B.        Bailey 4-1/4 acres        Not given
                                                              T. R. W.
  8. Canterbury   M. and B., O.    Whole area 3 acres        Risen.
     (Dungeon Hill)
  9. Carlisle     K. and B., O.    Whole area 4 acres         ...
 10. Chester      M. and B., O.    First ward 3/4 acre       Risen.
 11. Colchester   K. and B.       {Inner ward and keep}      Risen.
                                  {about 2 acres      }
 12. Dover        K. and B.       {Inner castle }            Risen.
                                  {about 6 acres}
 13. Durham       M. and B., O.    Bailey 1 acre              ...
 14. Ely          M. and B., O.    Bailey 2-1/2 acres        Fallen, but
                                                              rising.
 15. Exeter       B. only now      2 acres                    ...
 16. Gloucester   M. and B., O.            ?                 Risen.
 17. Hastings     M. and B., O.            ?                 Fallen, but
                                                              rising.
 18. Hereford     M. and B.        Bailey 5-1/2 acres        Risen.
 19. Huntingdon   M. and B., O.    Inner bailey 2-1/2 acres  Stationary.
 20. Lewes        M. and B.        Bailey 3 acres            Risen.
 21. Lincoln      M. and B.        Bailey 5-3/4 acres        Risen.
 22. Monmouth     K. and B.        Bailey 1-3/4 acres        Not given
                                                              T. R. E.
 23. Newcastle    M. and B., O.    Whole area 3 acres 1 rood  ...
 24. Norwich      M. and B., O.    Inner bailey 3-1/4 acres  Risen.
 25. Nottingham   M. and B., O.    Bailey 1-2/3 acres        Risen.
 26. Oxford       M. and B., O.    Bailey 3 acres            Risen.
 27. Pevensey     K. and B.        Bailey 1 acre             Risen.
 28. Quatford    {M. and B., }
                 {probably O.}     Bailey 1 acre              ...
 29. Rochester    M. and B., O.    Whole area about 3 acres  Risen.
     (Boley Hill)
 30. Old Sarum.   M. and B.        Inner ward 1-3/4 acres    Risen.
 31. Shrewsbury   M. and B., O.    Bailey 4/5 of an acre     Risen.
 32. Stafford     M. and B., O.    Bailey 1-3/5 acres        Risen.
 33. Stamford     M. and B.        Bailey 1-3/4 acres        Risen.
 34. Tamworth    {M. and B., }     Bailey 1 acre             Not given.
                 {probably O.}
 35. Totnes       M. and B., O.    Bailey 3/4 of an acre     Risen.
 36. Tower of     K. and B.         Originally?              Not given.
     London
 37. Wallingford  M. and B.        Bailey 4-1/2 acres        Risen.
 38. Warwick      M. and B., O.    Bailey 2-1/2 acres        Risen.
 39. Winchester   M. and B., O.    Whole area 4-1/2 acres    Not given.
 40. Worcester    M. and B., O.    Whole area between 3
                                     and 4 acres             Risen.
 41. York         M. and B., O.    Whole area formerly
                                     about 3 acres           Risen.
 42. The Baile    M. and B., O.    Whole area 2-3/4 acres    ...
     Hill, York


IN MANORS

                            Head of   Whole Area
 No. Name of                District  of Enceinte
     Castle.        Type.   T. R. E.  or Bailey.             Value.

 43. Abergavenny   M. and B.   ?     Bailey 1 acre.           ...
 44. Belvoir       M. and B.?  No         ?                  Risen.
 45. Berkeley      M. and B    Yes     "    1-1/2 acres      Risen.
     or Ness
 46. Berkhampstead M. and B    Yes     "    3 acres          Fallen.
 47. Bishop's      M. and B    No      "    2-1/2 acres      Fallen.
     Stortford
 48. Bourn         M. and B.   Yes     "    3 acres          Risen.
 49. Bramber       M. and B.   No      "    3 acres          Risen.
 50. Carisbrooke   M. and B.   No      "    2-3/4 acres      Risen.
 51. Castle Acre   M. and B.   No      "    2 acres          Risen.
 52. Chepstow      K. and B.   No    Whole area 1-2/3 acres  Risen.
 53. Clifford      M. and B.   No    Bailey 2-1/3 acres      Risen.
 54. Clitheroe     M. and B.   No      "    1 acre           Fallen.
 55. Corfe         M. and B.   No      "    1-1/2 acres      Risen.
 56. Dudley        M. and B.   No      "    1-3/4 acres      Fallen.
 57. Dunster       M. and B.   No      "    1-3/4 acres      Risen.
 58. Ewias         M. and B.    ?      "    2-1/3 acres      Not given
                                                              T. R. E.
 59. Eye           M. and B.   No      "    2 acres          Risen.
 60. Launceston    M. and B.   No      "    3 acres          Fallen.
 61. Montacute     M. and B.   No         ?                  Not given
                                                              T. R. E.
 62. Morpeth       M. and B.    ?         ?                   ...
 63. Norham        M. and B.    ?    Bailey 2 acres           ...
 64. Okehampton    M. and B.   No      "    1/2 an acre      Risen.
 65. Oswestry      M. and B.   No         ?                  Risen.
 66. Peak Castle   K. and B.   No      "    1 acre           Risen.
 67. Penwortham    M. and B.   No         ?                  Risen.
 68. Peterborough  Motte only   ?         ?                   ...
                     now
 69. Pontefract    M. and B. Probably  "    2-1/3 acres      Fallen.
 70. Preston Capes M. and B.   No          ...               Risen.
 71. Rayleigh      M. and B.   Yes   Bailey 3/4 acre         Risen.
 72. Richard's     M. and B.   No      "    2/3 acre         Stationary.
     Castle
 73. Richmond      K. and B.   No      "    2-1/2 acres       ...
 74. Rockingham    M. and B.   No    First bailey 3 acres    Risen.
 75. Skipsea       M. and B.   No    Bailey 8-1/4 acres      Fallen.
 76. Stanton       M. and B.   No         ?                  Risen.
     Holgate
 77. Tickhill      M. and B.   No      "    2 acres          Risen.
 78. Tonbridge     M. and B.   No      "    1-1/2 acres      Stationary.
 79. Trematon      M. and B.   No      "    1 acre           Fallen.
 80. Tutbury       M. and B.   No      "    2-1/2 acres      Not given
                                                              T. R. E.
 81. Tynemouth         ?        ?         ?                   ...
 82. Wigmore       M. and B.   No      "    1 acre           Risen.
 83. Windsor       M. and B.   No    Upper bailey            Fallen,
                                       6-1/2 acres           but rising.
 84. Wisbeach      M. and B.   No    Whole area 4 acres      Fallen.

It has been thought best to tabulate the _chief_ defensible area of
each castle. The total area, including ditches and scarps, is liable to
great variation owing to the nature of the ground.




INDEX


 Aber, 261

 Aberavon, 296

 Abercorn, 308

 Aberdovey, 300

 Abereinon, 301

 Abergavenny, 97

 Aberlleinog, 261

 Aberystwyth, 281, 393

 Aggeres, 77, 111

 Aldreth, 150

 Alfred, King, 13, 14, 15

 Amwell, 52

 Annan, 309

 Anstruther family, 308

 Antrim, 331

 Appledore, 50

 Aq'i, 331

 Aquila, castle of, 77

 Ardfinnan, 331

 Ardmayle, 331

 Ardnurcher, 331

 Ardree, 331

 Ardres, 75, =89=

 Area of Norman castles, 97

 Arques, 361

 Arundel, 98

 Arx, 211, 384

 Ashlar masonry, 356

 Askeaton, 332

 Askelon, 332

 Athelney, 14

 Athlone, 333

 Auchterless, 309

 Avenel family, 308


 Baginbun, 333

 Bailey, ballium, 4, 5, 92, 207

 Bakewell, 47

 Balimore Eustace, 333

 Balliol family, 308

 Ballyknockan, 341

 Ballynaclogh, 346

 Bamborough, 11, 100, 355, 357

 Banff, 319

 Barclay, 309

 Barnstaple, 102

 Barnwell, 376

 Baronies, 307

 Basements of keeps, 359, 362

 Basingwerk, 267

 Bastille, the, 376

 Bayeux Tapestry, 87, 158, 393

 Bayford Court, 49

 Bedford, 40

 Beith, 317

 Belesme, Roger, 100, 191;
   castle, 77

 Belvoir, 102

 Benfleet, 50

 Bensington, 28

 Berkeley, 103, 367

 Berkhampstead, 105

 Bernard de Neufmarché, 273, 276

 Bervie River, 316

 Biggar, 313

 Bishop's Stortford, 107

 Blaenporth, 282

 Bleddfa, 293

 Blois, 75

 Blythe, 219

 Boley Hill, 49, 196, 199, 200

 Bordlands, 307

 Borgue, 317

 Boroughs, 21, 258, 382

 Boulogne, 376 _n._ 1

 Bourn, 107

 Bowes, 366

 Bramber, 109

 Braose, De, 109, 276, 292

 Brecknock, 276, 290

 Bremesbyrig, 32

 Bretasche, 91, 386

 Bridgenorth, 33, 52

 Bristol, 23, 110

 Bromborough, 32

 Bruce family, 309

 Brut y Tywysogion, 254

 Buckingham, 25, 40

 Burghal Hidage, 28, 98, 160, =385=

 Burgh Castle, 44 _n._ 1.

 Burgus, 85

 Burh, 17-19, 123;
   Clark's theory of, 20-29

 Buttington, 51


 Cadwalader, 266

 Cadzow, 314

 Caen keep, 361

 Caereinion, 300

 Caerleon, 113

 Caerphilly, 376 _n._ 1

 Caerwedros, 283

 Caherconlish, 333

 Cambridge, 55, 57, =115=

 Camps, of refuge, 29;
   prehistoric, 6;
   of Danes, 61

 Canterbury, 116

 Carbury, 333

 Cardiff, 294

 Cardigan, 274, 275;
   Castle, =280=

 Carisbrook, 121

 Carlingford, 334

 Carlisle, 25, =128=, 365

 Carlovingian Empire, 66

 Carmarthen, =275=, 394

 Carnarvon, =261=, 375

 Carnwath, 318

 Carreghova, 271, 394

 Carrick, 334

 Carrickfergus, 334

 Carrickittle, 339

 Carrigogunell, 332

 Castel, the word, 24, 98

 Castellum, castrum, 25, 67, 169, 383

 Castles, private, Ch. V.;
   product of feudalism, 66;
   in Normandy, 76, 77;
   wooden, 78;
   stone, Ch. XII.;
   sites given to church, 259 _n._ 2.

 Castle Acre, =124=

 Castledermot, 347

 Castle guard, 175

 Castleknock, 341

 Castlemore, 338

 Castle Rough, 49

 Castle Rushen, 377 _n._ 1

 Castletown Delvin, 334

 Cathcart family, 310

 Catter, 315

 Ceredigion, 274, 275

 Chapels in castles, 355

 Chartres keep, 74

 Chastell Gwalter, 282

 Chateaudun keep, 75

 Chateau Gaillard, 372, 376, 391

 Chepstow, 125

 Chester, 31, =126=, 367

 Chevron moulding, 100

 Cheyne family, 310

 Chilham, 368

 Chimneys, 357

 Chinon, 75

 Chippenham, 55

 Chirk, 269

 Christison, Dr, 8, 31, 304, 306

 Cilgerran, =281=, 394

 Citadels, 6, 54, 56

 Clare, house of, 275, 281

 Clark, G. T., 2, 8, 19, 26, 48

 Clears, St, =288=, 394

 Cledemuthan, 43

 Clifford, 128

 Clitheroe, 129

 Clonard, 334

 Clonmacnoise, 335

 Colchester, 41, =132=, 223, 354, 355

 Collacht, 335

 Colville family, 310

 Comyn family, 310

 Concentric castles, 375

 Cooking in castles, 359

 Corfe, =135=, 363

 Coucy, 65, 372

 Courcy, John de, 336

 Court hills, 310, 391

 Covington, 313

 Crail, 319

 Crimond, 311

 Crogen, battle of, 266

 Cromarty, 316

 Crometh, 335

 Cromwell, 220, 336

 Croom, 335

 Crossbow, the 370

 Cunningham family, 311

 Cupar, 320

 Cymmer, 299

 Cynewulf, murder of, 13

 Cynfael, 300

 Cyricbyrig, 37


 Dalswinton, 311

 Dane John, 116, 118, 121

 Danes in Ireland, 60

 Dangio, 75

 Danish raths, 48;
   camps, 61;
   colonies, 59, 60

 Darnhall, 317

 David I. of Scotland, 123, 163, =303=

 Deganwy, 259, 270

 Dernio, 267

 Derver, 341

 Dinan, 87

 Dinerth, 282

 Dinevor, =287=, 394

 Dinweiler, 394

 Dirleton, 319

 Domfront keep, 361

 Donjons, 358, 390

 Douglas family, 312

 Dover, =138=;
   church, 144;
   Pharos, 143

 Downpatrick, 335

 Drogheda, 336

 Drumore, 315

 Drumsagard, 317

 Duchal, 315

 Dudley, 144

 Dudo of St Quentin, 76

 Duffus, 317

 Duleek, 337

 Dumfries, 320

 Dun, the word, 326

 Dunamase, 337

 Dungarvan, 337

 Dunio, 75

 Dunmullie, 311

 Dunoon, 313

 Dunskeath, 320

 Dunster, 145

 Durand, 312

 Durham, 146

 Durward, 312

 Dyfed, 274


 Earthworks, Committee, 2

 Eddisbury, 35

 Edward, 14-16, 45, 65, 127

 Edward the Martyr, 135

 Edwardian castles, 328, 345

 Egloe, Eulo, 271

 Elgin, 320

 Ellon, 311

 Ely, 149

 Entrances to keeps, 355, 361, 373

 Errol, 314

 Escluen, 332

 Étampes Castle, 373

 Ethelfleda, 14, 15, 16, 20, 21, 45, 232, 342

 Eu, 56

 Eustace of Boulogne, 139

 Ewias, 150

 Exeter, 151;
   siege of, 154

 Eye, 155


 Falaise, 361-363

 Favorie, 338

 Ferns, 338

 Feudalism, 63, 66, 378, 379;
   in Normandy, 76;
   in Wales, 299, 378;
   in Scotland, 303

 Fireplaces, 357

 Fitz Alans, 179, 313

 Fitzhardinge, Robert, 104

 FitzOsbern, William, 103, 126, 128, 150, 273

 Five Boroughs, the 44, 59

 Flambard, Ranulf, 172

 Fleming family, 313

 Flint, 375

 Folk-moots, 381

 Fore, 338

 Forebuildings, 355, 361

 Forfar, 320

 Forres, 321

 Fortifications, Anglo-Saxon, 29, 64;
   Danish, 55, 61;
   wooden, 78

 Fotheret Onolan, 338

 French earthworks, 7

 Fulham, 55

 Fulk Nerra, 73, 74, 352


 Gaimar, Geoffrey, 171

 Gallo-Roman villas, 67

 Galtrim, 338

 Gatehouse keeps, 359

 Gatehouse palace, 375

 Geashill, 338

 Gemaron, 292

 Gephthine, 332

 Gilling, 193, 194

 Gisors, 364

 Glamorgan, 276

 Gloucester, 156

 Godwin, Earl, 22, 24, 103

 Gomme, G. L., 8

 Gould, I. C, 2

 Gower, 277, 297

 Graham family, 314

 Granard, 338

 Greenwell, Canon, 2

 Guildford, 393

 Guisnes, 75, 76

 Gundulf, Bishop, 197, 198, 222

 Guy of Amiens, 139

 Gwyddgrug, 260

 Gwynedd, 256-262


 Hæsten the Dane, 49, 50

 Hall, the Anglo-Saxon, 17, 24, 168

 Hallaton Castle Hill, 88

 Hamilton family, 314

 Harold, Earl and King, 138, 161, 257

 Hastings, 87, =158=

 Haughead Kipp, 317

 Haverfordwest, 280

 Hawarden, 377

 Hawick, 315

 Hay, 291

 Hay family, 314

 Hên Domen, 395

 Henry I., castles of, 360-364, 392

 Henry II., castles of, 365-369

 Henry the Fowler, 64

 Hericio, 388

 Hermitage Castle, 318

 Herring-bone work, 136, 168, 218

 Hincheleder, 339

 Hithes, 57

 Hodesley, Hoseley, 268

 Holywell, 267, 270

 Hubert de Burgh, 140

 Hugh of Avranches, 256, 257

 Humphrey's Castle, 283

 Huntingdon, 42, =162=

 Hurdicia, 91, 372, =387=


 Ida, King, 11

 Inchelefyre, 339

 Innermessan, 321

 Inverness, 321

 Inverugie, 310

 Inverwick, 313

 Irish chiefs, 325, 342


 Jedburgh, 321

 John, Bishop of Terouenne, 88

 John, King, 137, 270, 370 _n._ 1

 Jomsborg, 59


 Karakitel, 339

 Keepless castles, 328, 374

 Keep and bailey, 357

 Keeps, arrangements in, 359, 391;
   polygonal, 368;
   prows to, 373;
   residences, 392;
   round, 368, 373

 Keeps of Henry I., 360, 363, 392

 Keeps of Henry II., 366

 Keeps of William I., 351, 354

 Kelts of Scotland, 304

 Kenardington, 50

 Kenfig, 295

 Kenmure, 308

 Kidwelly, 288

 Kilbixie, 339

 Kilbride, 318

 Kilfeakle, 340

 Kilfinnane, 329

 Kilkea, 347

 Killamlun, 339

 Killare, 339

 Kilmaurs, 311

 Kilmehal, 340

 Kilmore, 340

 Kilsantan, 340

 Kiltinan, 340

 Kincleven, 321

 Kirkcudbright, 321

 Kirkintilloch, 311

 Kirkpatrick Durham, 312

 Kitchens in castles, 90

 Knighton, 293

 Knock, 341

 Knockgraffan, 341


 Lacy, Ilbert de, 187, 188;
   Hugh de, 337

 Lag Castle, 312

 Lagelachon, 341

 Lagmen, 62

 Lambert of Ardres, 75, 89

 Lanark, 321

 Langeais keep, 72, 353

 Laon, 72

 Largs, 317

 Laugharne, 288

 Launceston, 164

 Laurence of Durham, 147

 Law about castles, 377

 Lawhaden, 279

 Lea Castle, 341

 Lea River, 15, 52

 Lead roofs, 369

 Leighlin, 341

 Lennox, 315

 Leuchars, 318

 Lewes, 165

 Lincoln, 167

 Linton Roderick, 310

 Lismore, 342

 Llanarmon, 272

 Llandeilo Talybont, 298

 Llandovery, 286

 Llanegwad, 289

 Llangadog, 289, 394

 Llanrhystyd, 300

 Llanstephan, 286

 Lloyd, Professor, 253, 393

 Lochmaben, 309

 Lochorworth, 314

 Lockhart family, 315

 Logan family, 315

 London fortified, 14

 Loopholes, 362, 370, 371;
   cross loopholes, 371

 Lords-marchers, 255

 Loske, 343

 Loughor, 298

 Louth, 342

 Louvre, the, 373

 Lovel family, 315

 Loxhindy, 343

 Ludgarsburh, 170

 Lumphanan, 312

 Lyle or Lisle family, 315

 Lympne, 15


 Mabudryd, 394

 Machicolations, 372

 Magh Adair, 327

 Maitland, Professor, 27

 Maldon, 41

 Manchester, 46

 Manors, Saxon, and mottes, 96

 Mans, Le, keep of, 361

 Masonry, 356, 362, 367

 Mathraval, 271

 Maud's Castle, 293, 394

 Maxton, 316

 Maxwell family, 316

 Melton, Archbishop, 248

 Melville family, 315

 Mercenaries, 7, 74 _n._ 1

 Merchem Castle, 88

 Mersey Island, 54

 Military service, 64

 Milton, 49

 Missile engines, 369

 Mitford, 373

 Moffat, 309

 Mold, 260

 Monmouth, 168

 Montacute, 169

 Montalt, 316

 Montgomeri, Roger de, 53, 98, =130=, 191, 263;
   Hugh de, 274;
   Arnolf, 274, 278;
   castle, =264=, 395

 Montgomerie family, Scotland, 316

 Moot-hills, 8, 9, 381

 Moray, colonisation of, 304

 Morpeth, 171

 Mortain, Count of, 106, 138, 164, 169, 186

 Mortimers, 276

 Morville family, 316

 Mottes, described, 4, 5;
   the word, 9 _n._ 1;
   distribution, 80-82;
   situation, 83-96;
   in France, 85;
   in Wales, 301;
   in Scotland, 322;
   in Ireland, 348;
   history, 72, 74

 Mowbray, Earl Robert, 101

 Mowbray family, 317

 Müller, Dr Sophus, 6

 Mural towers, 358

 Murray family, 317


 Naas, 343

 Nantes Castle, 71

 Nant yr Arian, 284

 Narberth, 280

 Navan, 344

 Neath, 296

 Neckham, "De Utensilibus," 86

 Neilson, Mr George, 8, =306=

 Neu Leiningen, 376 _n._ 1

 Newcastle, 171

 Newcastle Bridgend, 295

 Newcastle Emlyn, 289

 New Grange, 10

 Newport, 279

 Nicetus, his castrum, 67

 Nicholas, St, 110, 213

 Nobber, 344

 Normandy, 22, 76, 77

 Norman favourites, 23

 Norman walls, 356

 Normans, 7

 Norrei Castle, 77

 Northmen, camps of, 61, 383

 Norton, 293

 Norwich, =173=, 363

 Nottingham, 44, 49, 57, =176=


 O'Donovan, 329

 Offa's Dyke, 52

 Okehampton, 178

 Oldcastle, 347

 Old Sarum, 202

 Oliphant family, 317

 Orford, 246 _n._

 Oswestry, 179

 Overton (Denbigh), 267;
   (Hereford), 192

 Owen Gwynedd, 260, 270

 Oxford, 180

 Oxnam, 310

 Oystermouth, 298


 Pantolf, William, 213

 Parliamentary fortifications, 202

 Payn's Castle, =293=, 395

 Peak, 182

 Pembroke, 278

 Pentecost's Castle, 24, 150

 Penwortham, 183

 Peterborough, 185

 Pevensey, 99, 186

 Pistes, Capitulary of, 68, 72

 Pitt-Rivers, General, 2

 Plinths, 355

 Polnoon, 316

 Pontefract, 187;
   siege of, 189

 Pont y Stuffan, 283

 Powys, 263-266

 Prestatyn, 270

 Preston Capes, 190

 Pretorium, 72

 Prisons in castles, 359

 Private castles, 21, 68

 Pudsey, Bishop, 173


 Quatbridge, 58

 Quatford, 191

 Quincy, De, family, 318


 Radnor, 292

 Rainald the Sheriff, 263

 Rapes of Sussex, 299

 Rathceltchair, 336

 Raths in Ireland, 325, 327

 Rathwire, 344

 Ratouth, 344

 Rayleigh, 191

 Reading, 54

 Redcastle, Lunan Bay, 309

 Reginald's Tower, 347, 348

 Remni, 297

 Renfrew, 313

 Retford, 55

 Rhaidr Gwy, 301

 Rhé Island, motte on, 350

 Rhuddlan, 257, 259

 Rhyd y Gors, 275, 284

 Rhys ap Griffith, 275, 287

 Riccarton, 319

 Richard Sans Peur, 76

 Richard I., 370-372

 Richard's Castle, 192

 Richmond, 193

 Robert Curthose, 110

 Robert de Monte, 360

 Robert, Earl of Gloucester, 110, 295

 Robert Fitz Hamon, 277

 Robert of Rhuddlan, 257

 Roberton, 314

 Rochester, 25, 49, 195

 Rockingham, 201

 Roger the Poitevin, 129, 183, 184

 Rokerel, 345

 Rollo, 76

 Roscrea, 345

 Rosemarkie, 322

 Ross, 318

 Rouen, 77

 Runcorn, 38

 Ruthin, 269

 Ruthven, 311


 Sanquhar, 318

 Sarn Helen, 283

 Saumur Castle, 75

 Saxon fortifications, Chapters II., III., 29

 Saxon royal seats, 151 _n._ 2, 235

 Scergeat, 33

 Sepulchral hillocks, 9;
   in Ireland, 327

 Shaftesbury, 15

 Shell keep, 99

 Sheppey Isle, 54

 Shoebury, 51

 Shrewsbury, 207

 Siege castles, 85

 Siegfried the Dane, 75

 Siward, Earl, 22

 Skipsea, 209

 Skreen, 345

 Slane, 345

 Somerville family, 318

 Somner, antiquary, 117

 Soulis family, 318

 Stafford, 34, =211=

 Stamford, 44, =216=

 Stanton, 217

 Stevenston, 315

 Stewarton, 318

 Swansea, 297

 Symington, 315


 Table of Boroughs, 26

 Talgarth, 292

 Tamworth, 34, =218=

 Tarbolton, 314

 Tateshall, 187

 Tempsford, 53

 Tenby, 280

 Terraces to mottes, 102

 Thanet, 54

 Thelwall, 46, 385

 Thetford, 55, 56

 Thibault-le-Tricheur, 74

 Thingwall, 381

 Thorne, 368 _n._ 3

 Thurles, 346

 Tibraghny, 346

 Tickhill, 219

 Tiles, use of, 255

 Timahoe, 346

 Tom-a-mhoid, 313

 Tomen y Mur, 262

 Tomen y Rhoddwy, 271, 272, 388

 Tonbridge, 220

 Toot Hill, 259

 Topcliffe, 5 _n._

 Torkesey, 55, 56

 Totnes, 221

 Towcester, 41

 Tower of London, 221, 354, 355

 Towers to castles, 71

 Towns, fortification of, 65

 Trade, 30

 Trébuchet, 369

 Trematon, 226

 Tribalism, 64

 Trim, 346, 377

 Tristerdermot, 347

 Tullow, 335, 389

 Tutbury, 227

 Tynboeth, 293

 Tynemouth, 228

 Tynwald Hill, 381

 Typermesan, 347


 Valoignes family, 318

 Value of manors and towns, 96

 Vaulting, 362, 365, 367, 374

 Vaux family, 319

 Viking crews, 90

 Viollet le Duc, 368 _n._ 1

 Vire, keep, 361

 Voussoirs, 362


 Wales, Chapters VIII., IX.;
   Wales and Saxons, 253;
   Wales and Normans, 254

 Wallace family, 319

 Wallingford, 28, =228=

 Walwern, 301

 Wareham, 25, 28

 Warenne, Wm., 124, 165

 Wark, 366, 367, 368 _n._ 2, 388

 Warkworth, 377

 Warwick, 36, =230=

 Wasta, 114

 Waterford, 347

 Water-supply, 362-363

 Watling Street, 16, 32, 382

 Waytemore Castle, 107

 Weardbyrig, 36

 Welsh halls, 251

 Welshpool, 265

 Wessex, 13

 Wexford, 348

 Wicklow, 348

 Wigingamere, 41

 Wigmore, 232

 William I., 22, 77

 William the Lion, 163, 305

 Willington, 57

 Winchester, 233

 Winding walks on mottes, 102, 121

 Windsor, 236;
   borough, 238

 Wisbeach, 239

 Wiston, 279

 Witham, 39

 Wolvesey Castle, 235 _n._ 4

 Wooden fortifications, 78, 208, 228, 250, 306, 358, 359

 Worcester, 115, 23, 31, =240=;
   charter, 21

 Wrexham Castle, 268


 Yale Castle, 271, 300

 Year 1000, 78

 York, 13, =242=

 York, Baile Hill, 248

 Ystrad Cyngen, 288

 Ystrad Meurig, 283

 Ystrad Peithyll, 282


PRINTED BY OLIVER AND BOYD, EDINBURGH FOOTNOTES:

[1] Mr W. H. St John Hope arrived independently at similar conclusions.

[2] In the paper on Earthworks in the second volume of the _Victoria
County History of Yorkshire_, this subdivision of the promiscuous class
X., is used.

[3] Since the above was written, Mr Hadrian Allcroft's work on
_Earthwork of England_ has furnished an admirable text-book of this
subject.

[4] See Frontispiece.

[5] See Fig. 1.

[6] For instance, at Berkeley, Ewias Harold, Yelden, and Tomen y Roddwy.

[7] As at Rayleigh and Downpatrick.

[8] In some of these castles there is no gap in the bailey banks for an
entrance. They must have been entered by a movable wooden stair, such
as horses can be taught to climb. See the plan of Topcliffe Castle,
Yorks (Fig. 1).

[9] _Vor Oldtid_, p. 629.

[10] _Entwickelung des Kriegswesens_, iii., 379.

[11] See Chapter VII.

[12] _Primitive Folkmoots._ See Appendix A.

[13] _Early Fortifications in Scotland_, p. 13. He adds an instance
showing that Moot Hill is sometimes a mistake for Moot Hall.

[14] _Scottish Review_, vol. xxxii.

[15] Some writers give the name of moot-hill to places in Yorkshire
and elsewhere where the older ordnance maps give moat-hill. _Moat_ in
this connection is the same as _motte_, the Scotch and Irish _mote_,
i.e., the hillock of a castle, derived from the Norman-French word
_motte_. As this word is by far the most convenient name to give to
these hillocks, being the only specific name which they have ever had,
we shall henceforth use it in these pages. We prefer it to _mote_,
which is the Anglicised form of the word, because of its confusion
with _moat_, a ditch. Some writers advocate the word _mount_, but this
appears to us too vague. As the word _motte_ is French in origin, it
appropriately describes a thing which was very un-English when first
introduced here.

[16] At York, a prehistoric crouching skeleton was found by Messrs
Benson and Platnauer when excavating the castle hill in 1903, 4 feet
6 inches below the level of the ground. The motte at York appears
to have been raised after the destruction of the first castle, but
whether the first hillock belonged to the ancient burial is not decided
by the account, "Notes on Clifford's Tower," by the above authors.
_Trans. York. Philosoph. Soc._, 1902. Another instance is recorded in
the _Revue Archæologique_, to which we have unfortunately lost the
reference.

[17] From the report of a competent witness, Mr Basil Stallybrass.

[18] Earle, _Two Saxon Chronicles Parallel_, Introd., xxiii.

[19] Nennius says that Ida "_unxit_ (read cinxit) Dynguayrdi
Guerth-Berneich"=a strength or fort of Bernicia. _Mon. Hist. Brit._,
75. Elsewhere he calls Bamborough Dinguo Aroy. It is quite possible
that there might have been a Keltic _din_ in a place so well fitted for
one as Bamborough.

[20] Bede, H. E., iii., 16.

[21] See Bede, as above, and Symeon, ii., 45 (R.S.).

[22] We infer this from the strong defences of what is now the middle
ward.

[23] The fact, however, that the _Trinoda Necessitas_, the duty of
landholders to contribute to the repair of boroughs and bridges, and to
serve in the fyrd, is occasionally mentioned in charters earlier than
the Danish wars, shows that there were town walls to be kept up even at
that date. See Baldwin Brown, _The Arts in Early England_, i., 82.

[24] See Wright, _History of Domestic Manners_, p. 13.

[25] The Danish fortress of Nottingham is mentioned by the _Chronicle_
in 868, but we are speaking now of purely Anglo-Saxon fortresses.

[26] Asser, ch. 91, Stevenson's edition.

[27] "That same year King Alfred repaired London; and all the English
submitted to him, except those who were under the bondage of the
Danish men; and then he committed the city (_burh_) to the keeping of
Ethelred the ealdorman." _A.-S. C._, 886. The word used for London is
_Londonburh_. Asser says: "Londoniam civitatem honorifice restauravit
et habitabilem fecit," p. 489.

[28] Birch's _Cartularium_, ii., 220, 221.

[29] _Anglo-Saxon Chronicle_, 878, 893, 896. According to Henry of
Huntingdon, the work on the Lea was the splitting of that river into
two channels; but I am informed that no trace of such a division
remains.

[30] _Gesta Pontificum_, 186. See Appendix C.

[31] Birch's _Cartularium_, ii., 222; Kemble's _Codex Diplomaticus_,
v., 142.

[32] He signs a charter in 889 as "subregulus et patricius Merciorum,"
Kemble's _Codex Diplomaticus_. See Freeman, _N. C._, i., 564; and
Plummer, _A.-S. C._, i., 118.

[33] The dates in this chapter are taken from Florence of Worcester,
who is generally believed to have used a more correct copy of the
_Anglo-Saxon Chronicle_ than those which have come down to us.

[34] See Appendix B.

[35] _A.-S. C._, 910, 911.

[36] _New English Dictionary_, Borough.

[37] _Anglo-Saxon Chronicle_, 942. The Anglo-Saxon Chronicle has three
words for fortifications, _burh_, _faesten_, and _geweorc_. Burh is
always used for those of Edward and Ethelfleda, faesten (fastness) or
geweorc (work) for those of the Danes.

[38] See the illustrations in Wright, _History of Domestic Manners_.

[39] _Bury_ is formed from _byrig_, the dative of _burh_.

[40] Professor Maitland observed: "To say nothing of hamlets, we
have full 250 parishes whose names end in burgh, bury, or borough,
and in many cases we see no sign in them of an ancient camp or of an
exceptionally dense population." _Domesday Book and Beyond_, 184.

[41] Schmid, _Gesetze der Angelsachsen_, pp. 176, 214, 372. It is not
absolutely certain that the _burh_ in these three cases does not mean a
town.

[42] Schmid, 138. Professor Maitland says: "In Athelstan's day it seems
to be supposed by the legislator that a _moot_ will usually be held in
a _burh_. If a man neglect three summonses to a moot, the oldest men
of the _burh_ are to ride to his place and seize his goods." _Domesday
Book and Beyond_, 185. "All my reeves," are mentioned in the Preface to
_Athelstan's Laws_, Schmid, 126.

[43] Schmid, 138. "Butan porte" is the Saxon expression, _port_ being
another word for town; see Schmid, 643.

[44] Schmid, Edgar III., 5; Ethelred II., 6.

[45] Edgar IV., 2.

[46] The writer was first led to doubt the correctness of the late Mr
G. T. Clark's theory of burhs by examining the A.-S. illustrated MSS.
in the British Museum. On p. 29 of the MS. of _Prudentius_ (Cleopatra,
c. viii.), there is an excellent drawing of a four-sided enclosure,
with towers at the angles, and battlemented walls of masonry. The
title of the picture is "Virtutes urbem ingrediuntur," and _urbem_ is
rendered in the A.-S. gloss as _burh_. See Fig. 2.

[47] Florence translates _burh_ as _urbs_ nineteen times, as _arx_ four
times, as _murum_ once, as _munitio_ once, as _civitas_ once.

[48] Published in 1884, but comprising a number of papers read to
various archæological societies through many previous years, during
which Mr Clark's reputation as an archæologist appears to have been
made.

[49] "Eallum thæm folc to gebeorge." Birch's _Cartularium_, ii., 222.

[50] Professor Maitland has claimed that the origin of the boroughs
was largely military, the duty of maintaining the walls of the county
borough being incumbent on the magnates of the shire. _Domesday Book
and Beyond_, 189. See Appendix C.

[51] Parker's _Domestic Architecture in England from Richard II. to
Henry VIII._, part ii., 256.

[52] _A.-S. C._, 1048.

[53] _William of Jumièges_, vii.-xvii.

[54] _A.-S. C._ (Peterborough), 1048.

[55] _A.-S. C._, 1052 (Worcester). This castle is generally supposed
to be Richard's Castle, Herefordshire, built by Richard Scrob; but I
see no reason why it should not be Hereford, as the Norman Ralph, King
Edward's nephew, was Earl of Hereford. We shall return to these castles
later.

[56] Mr Freeman says: "In the eleventh century, the word _castel_ was
introduced into our language to mark something which was evidently
quite distinct from the familiar _burh_ of ancient times.... Ordericus
speaks of the thing and its name as something distinctly French:
"munitiones quas Galli castella nuncupant." The castles which were now
introduced into England seem to have been new inventions in Normandy
itself. William of Jumièges distinctly makes the building of castles to
have been one of the main signs and causes of the general disorder of
the days of William's minority, and he seems to speak of the practice
as something new." _N. C._, ii., 606. It is surprising that after so
clear a statement as this, Mr Freeman should have fallen under the
influence of Mr Clark's _burh_ theory, and should completely have
confused castles and boroughs.

[57] _Codex Diplomaticus_, i., 138.

[58] _History of Rochester_, 1772, p. 21.

[59] Stevenson's edition of _Asser_, 331. See Appendix D.

[60] _Asser_, c. xlix.

[61] Worcester, Chester, Tamworth, Stafford, Warwick, Hertford,
Buckingham, Bedford, Maldon, Huntingdon, Colchester, Stamford, and
Nottingham.

[62] _Domesday Book and Beyond_, 216.

[63] Buckingham is the only place which is included in both lists. See
Appendix E.

[64] _Domesday Book and Beyond_, 188. See Appendix E. Southwark, one of
the names, which is not called a borough in Domesday, retains its name
of _The Borough_ to the present day.

[65] No Roman remains have been found in either place.

[66] _Beauties of England and Wales_, Oxfordshire.

[67] See Skeat's _Dictionary_, "Timber."

[68] Excavation has recently shown that many of the great hill-forts
were permanently inhabited, and it is now considered improbable that
they were originally built as camps of refuge. It seems more likely
that this use, of which there are undoubted instances in historic times
(see Cæsar, _Bello Gallico_, vi., 10, and v., 21), belonged to a more
advanced stage of development, when population had moved down into the
lower and cultivatable lands, but still used their old forts in cases
of emergency.

[69] _Ante_, p. 21.

[70] Haverfield, in V. C. H. Worcester, _Romano-British Worcester_, i.

[71] _Early Fortifications in Scotland_, p. 105.

[72] Gairdner and Mullinger, _Introduction to the Study of English
History_, 268.

[73] The tower called Cæsar's Tower is really a mural tower of the 13th
century. E. W. Cox, "Chester Castle," in _Chester Hist. and Archæol.
Soc._, v., 239.

[74] Cox, as above. See also Shrubsole, "The Age of the City Walls of
Chester," _Arch. Journ._, xliv., 1887. The present wall, which includes
the castle, is an extension probably not earlier than James I.'s reign.

[75] The charter is given in Ormerod's _History of Cheshire_, ii., 405.

[76] _Journ. of Brit. Arch. Ass._, 1875, p. 153.

[77] _Itin._, ii., 2.

[78] "Arcem quam in occidentali Sabrinæ fluminis plaga, in loco qui
Bricge dicitur lingua Saxonica, Ægelfleda Merciorum domina quondam
construxerat, fratre suo Edwardo seniore regnante, Comes Rodbertus
contra regem Henricum, muro lato et alto, summoque restaurare coepit."
1101.

[79] A good deal has been made of the name Oldbury, as pointing to the
_old burh_; but Oldbury is the name of the manor, not of the hillock,
which bears the singular name of Pampudding Hill. Tradition says that
the Parliamentary forces used it for their guns in 1646. Eyton's
_Shropshire_, i., 132.

[80] "Bricge cum exercitu pene totius Angliæ obsedit, machinas quoque
ibi construere et castellum firmare præcepit." _Florence_, 1102.

[81] Florence in fact says _urbem restauravit_.

[82] D. B., i., 246.

[83] These buildings formed part of a hunting lodge built in the
reign of Edward III., called The Chamber in the Forest. See Ormerod's
_Cheshire_, ii., 3. When visiting Eddisbury several years ago, the
writer noticed several Perpendicular buttresses in these ruins.

[84] D. B., i., 238a, 1.

[85] "Abbas de Couentreu habet 36 masuras, et 4 sunt wastæ propter
situm castelli.... Hae masurae pertinent ad terras quas ipsi barones
tenent extra burgum, et ibi appreciatae sunt." D. B., i., 238.

[86] _Domesday Book and Beyond_, p. 189. See Appendix D.

[87] Dugdale's _Warwickshire_, 1st edition, pp. 50 and 75. The
derivation of Kirby from Cyricbyrig is not according to etymological
rules, but there can be no doubt about it as a fact; for in Domesday it
is stated that _Chircheberie_ was held by Geoffrey de Wirche, and that
the monks of St Nicholas [at Angers] had two carucates in the manor. In
the charter in which Geoffrey de Wirche makes this gift Chircheberie is
called Kirkeberia [_M. A._, vi., 996], but in the subsequent charter of
Roger de Mowbray, confirming the gift, it is called Kirkeby.

[88] _Britannia_, ii., 375.

[89] _Numismatic Chronicle_, 3rd S., xiii., 220.

[90] Fowler's _History of Runcorn_ gives a plan of this fort, and there
is another in Hanshall's _History of Cheshire_, p. 418 (1817). A very
different one is given in Beaumont's _History of Halton_.

[91] Beaumont's _Records of the Honour of Halton_. In 1368, John Hank
received the surrender of a house near to the castle in Runcorn.

[92] _Mediæval Military Architecture_, ii., 120.

[93] _Essex Naturalist_, January 1887.

[94] Danbury Camp, which has also been surveyed by Mr Spurrell (_Essex
Naturalist_, 1890), is precisely similar in plan to Witham, but nothing
is known of its history.

[95] See _Victoria History of Bedfordshire_, i., 281.

[96] Morant's _History of Essex_, i. Three sides of the rampart were
visible in his time.

[97] D. B., ii., 5.

[98] _Itin._, i., 12.

[99] Baker's _History of Northampton_, ii., 321.

[100] D. B., i., 219b.

[101] _A.-S. C._, 921. "Wrohte tha burg æt Tofeceastre mid stan
wealle." Florence says 918.

[102] Baker, _History of Northants_, ii., 318. See also Haverfield, _V.
C. H._, Northants, i., 184.

[103] Atkinson's _Cambridge Described_, p. 1.

[104] There is, however, this difficulty, that Cambridge was still
occupied by a Danish force when Wigingamere was built. It submitted to
Edward in 918.

[105] See Mr Plummer's discussion of these variations in his edition of
the _Chronicle_, ii., 116.

[106] Lewis, _Topographical Dictionary of England_. Mr Rye
remarks:--"The silting up of the harbour has ruined a port which
once promised to be of as great importance as Norwich." _History of
Norfolk_, p. 228.

[107] It is really wonderful that the identification of Cledemuthan
with the mouth of the Cleddy in Pembrokeshire could ever have been
accepted by any sober historian. That Edward, whose whole time was
fully occupied with his conquests from the Danish settlers, could have
suddenly transported his forces into one of the remotest corners of
Wales, would have been a feat worthy of the coming days of air-ships.
William of Worcester has preserved a tradition that Edward repaired
Burgh, "quae olim Saxonice dicebatur Burgh-chester," but he confuses it
with Norwich. _Itinerarium_, 337. Is it possible that we ought to look
for Cledemuthan at Burgh Castle, at the mouth of the Waveney? It would
be quite in accordance with Edward's actions elsewhere to restore an
old Roman _castrum_.

[108] Leland says: "There were 7 principall Towers or Wards in the
waulles of Staunford, to eche of which were certeyne freeholders in the
Towne allottid to wache and ward in tyme of neadde." _Itinerarium_,
vii., 11.

[109] _A.-S. C._, 868.

[110] Shipman's _Old Town Wall of Nottingham_, pp. 73-75. The evidence
for a Roman origin of the borough is altogether too slight, as, except
some doubtful earthenware bottles, no Roman remains have been found at
Nottingham.

[111] _A.-S. C._, 921. _Florence of Worcester_, 919.

[112] I am indebted for much of the information given here to the local
antiquarian knowledge of Mr Harold Sands, F.S.A. He states that the
old borough was 1400 yards from the Trent at its nearest point, and
that the highest ground on the south side of the Trent is marked by the
Trent Bridge cricket ground, the last spot to become flooded. Here,
therefore, was the probable site of Edward's second borough.

[113] See Appendix F.

[114] Whitaker's _History of Manchester_, i., 43.

[115] _Trans. of Lanc. and Chesh. Hist. and Ant. Soc._, v., 246.

[116] "Castle" in combination with some other word is often given to
works of Roman or British origin, because its original meaning was a
fortified enclosure; but the name Castle Hill is extremely common for
mottes.

[117] We may remark here that it is not surprising that there should
be a number of motte castles which are never mentioned in history,
especially as it is certain that all the "adulterine" castles, which
were raised without royal permission in the rebellions of Stephen's and
other reigns, were very short-lived.

[118] _Mediæval Military Architecture_, i., 18. See Mr Round's
remarks on Mr Clark's vagueness in his "Castles of the Conquest,"
_Archæologia_, 1902.

[119] The _A.-S. C._ speaks of this Danish host as "a great heathen
army." 866.

[120] "Worhton other fæsten ymb hie selfe." The same language is
frequently used in the continental accounts of the Danish fortresses:
"Munientes se per gyrum avulsæ terræ aggere," _Dudo_, 155 (Duchesne):
"Se ex illis (sepibus et parietibus) _circumdando_ munierant." _It._,
p. 81.

[121] The earthworks at Bayford Court must belong to the mediæval
castle which existed there. See _Beauties of England and Wales, Kent_,
p. 698. Castle Rough is less than an acre in area.

[122] Mr Harold Sands, _Some Kentish Castles_, p. 10.

[123] See the plan in _Victoria History of Kent_, paper on Earthworks
by the late Mr I. C. Gould. Hasted states that there was a small
circular mount there as well as an embankment, and that there are other
remains in the marsh below, which seem to have been connected with the
former by a narrow ridge or causeway, _Kent_, iii., 117. The causeway
led to a similar mount in the marsh below, but Mr Gould inclined to
think the mounts and causeway later, and possibly part of a dam for
"inning" the marsh. _V. C. H._, p. 397.

[124] "Hæsten's Camps at Shoebury and Benfleet," _Essex Naturalist_,
iv., 153.

[125] The _Chronicle_ says that the ships of Hæsten were either broken
to pieces, or burnt, or taken to London or Rochester. 894.

[126] _Essex Naturalist_, as above, p. 151. These berms certainly
suggest Roman influence.

[127] _A.-S. C._, 894.

[128] _Montgomery Collections_, xxxi., 337; Dymond, _On the Site of
Buttington_. See also Steenstrup, _Normannerne_, ii., 80.

[129] _Beauties of England and Wales_, vii., 246. There is nothing left
either at Great or Little Amwell now but fragments of what are supposed
to be homestead moats. _Royal Commission on Historical Monuments_, pp.
95, 142, Herts. vol.

[130] Florence's date.

[131] _Victoria History of Bedfordshire_, i., 282, from which this
description is taken.

[132] The _Chronicle_ speaks of _Tempsford_ as a _burh_, so it must
have been a large enclosure.

[133] Mr Clark actually speaks of a subsequent Norman castle at
Tempsford (_M. M. A._, i., 78), but we have been unable to find any
confirmation of this. Faint traces of larger works in the fields below
were formerly visible. _V. C. H. Bedfordshire_.

[134] Stephenson's _Asser_, p. 27.

[135] There are no remains of earthworks in Thanet or Sheppey, except a
place called Cheeseman's Camp, near Minster in Thanet, which the late
Mr Gould regarded as of the "homestead-moat type." _V. C. H. Kent_,
i., 433. Nor are there any earthworks on Mersey Island mentioned by Mr
Gould in his paper on Essex earthworks in the _V. C. H._

[136] Stukeley, who saw this earthwork when it was in a much more
perfect state, says that it contained 30 acres. See Mr Hope's paper in
_Camb. Antiq. Soc._, vol. xi.

[137] Blomefield's _Norfolk_, ii., pp. 7, 8, 27. His description is
very confused.

[138] See Erlingssen's _Ruins of the Saga Time, Viking Club_, p. 337.

[139] Richerii, _Historiarum Libri Quatuor_, edition Guadet, p. 67.

[140] "In modo castri, munientes se per girum avulsæ terræ aggere."
_Dudo_, 155 (edition Duchesne).

[141] "The castle end of Cambridge was called the Borough within the
memory of persons now living." Atkinson's _Cambridge Described_ (1897),
p. 9.

[142] Steenstrup says that the Northmen built themselves shipyards all
round Europe, especially on the islands where they had their winter
settlements. _Normannerne_, i., 354.

[143] A.-S., _hyth_, a shore, a landing-place.

[144] _Victoria County History of Beds._, i., 282.

[145] Steenstrup's _Normannerne_, vol. iv.; _Danelag_, p. 40.

[146] _A.-S. C._, 914-921.

[147] Steenstrup, _Danelag_, p. 41.

[148] _Ibid._, pp. 22, 23.

[149] Such quartering must have been confined to the unmarried Danes,
but there must have been plenty of unmarried men in the piratical host,
even at the period when it became customary to bring wives and children
with the army.

[150] _Normannerne_, i., 282.

[151] _Dudo_, 76 (Duchesne).

[152] Herr Steenstrup shows that so far from the settlement of the
Danes in Normandy being on feudal lines, they only reluctantly accepted
the feudal yoke, and not till the next century. _Normannerne_, i., 305,
310. It is not till the 11th century that feudal castles become general
in Normandy.

[153] The Danes in Normandy soon made Rouen a great centre of trade.
_Normannerne_, i., 190.

[154] Cunningham's _Growth of English Industry_, i., 92.

[155] See Vinogradoff, _English Society in the 11th Century_, pp. 5,
11, 478.

[156] See Stubbs, _Constitutional History_, i., 251; Maitland's
_Domesday Book and Beyond_, p. 157; Round's _Feudal England_, p. 261;
Vinogradoff's _English Society in the 11th Century_, p. 41.

[157] Professor Maitland wrote: "The definitely feudal idea that
military service is the tenant's return for the gift of land did not
exist [before the Norman Conquest], though a state of things had
been evolved which for many practical purposes was indistinguishable
from the system of knight's fees." _Domesday Book and Beyond_, p.
157. Dr Round holds that "the military service of the Anglo-Norman
tenant-in-chief was in no way derived or developed from that of the
Anglo-Saxons, but was arbitrarily fixed by the king, from whom he
received his fief." _Feudal England_, p. 261. Similarly, Professor
Vinogradoff states that "the law of military fees is in substance
French law brought over to England by the [Norman] conquerors."
_English Society in the 11th Century_, p. 41.

[158] Giesebrecht, _Geschichte der Kaiserzeit_, i., 224. The word
_Burg_, which Giesebrecht uses for these strongholds, means a castle
in modern German; but its ancient meaning was a town (see Hilprecht's
_German Dictionary_), and it corresponded exactly to the Anglo-Saxon
_burh_. It was used in this sense at least as late as the end of the
12th century; see, _e.g._, Lamprecht's _Alexanderlied_, passim. It is
clear by the context that Giesebrecht employs it in its ancient sense.

[159] _Ibid._, 222. Henry's son Otto married a daughter of Edward the
Elder. Henry received the nickname of Townfounder (Städtegründer).

[160] "Carolus civitates Transsequanas ab incolis firmari rogavit,
Cinomannis scilicet et Turonis, ut præsidio contra Nortmannos _populis_
esse possent." _Annales Bertinianorum_, Migne, Pat., 125, 53.

[161] Flodoard, _Hist. Ecc. Remensis_, iv., viii.

[162] Modern historians generally say that he built the _castle_ of
Coucy; but from Flodoard's account it seems very doubtful whether
anything but the town is meant. _Annales_, iv., xiii. His words
are: "Munitionem quoque apud Codiciacum tuto loco constituit atque
firmavit." _Munitio_ properly means a bulwark or wall.

[163] _Gesta Episcop. Cameracensium_, Pertz, vii., 424.

[164] _Chron. Camarense et Atrebatorum_, Bouquet, x., 196.

[165] Sismondi, _Histoire des Français_, ii., 172.

[166] Guizot, _Histoire de la Civilisation en France_, iii., 311.

[167] Enlart, _Manuel d'Archæologie Française_, ii., 494.

[168] See Dr Haverfield's articles in the _Victoria County Histories_,
passim. The late J. H. Burton justly wrote: "We have nothing from the
Romans answering to the feudal stronghold or castle, no vestige of a
place where a great man lived apart with his family and his servants,
ruling over dependants and fortifying himself against enemies."
_History of Scotland_, i., 385.

[169] _Annals of Fulda_, 394, Pertz, i.

[170] _Cap. Regum Francor._, ii., 360.

[171] Thus De Caumont unfortunately spoke of the fortress built
by Nicetus, Bishop of Treves, in the 6th century, as a _château_
(_Abécédaire_, ii., 382); but Venantius Fortunatus, in his descriptive
poem, tells us that it was a vast enclosure with no less than thirty
towers, built by the good pastor for the protection of his flock. It
even contained fields and vineyards, and altogether was as different
from a private castle as anything can well be. Similarly the _castrum_
of Merliac, spoken of by Enlart (_Architecture Militaire_, p. 492) as
a "veritable château," is described as containing cultivated lands
and sheets of water! (Cited from Gregory of Tours, _Hist. Francorum_,
liii., 13.) De Caumont himself says: "Les grandes exploitations rurales
que possédaient les rois de France et les principaux du royaume du
V^ième au X^ième siècle ne furent pas des forteresses et ne doivent
point être confondues avec les chateaux." _Abécédaire_, ii., 62.

[172] See Appendix D.

[173] "Volumus et expresse mandamus, ut quicunque istis temporibus
castella et firmitates et haias sine nostro verbo fecerint, Kalendis
Augusti omnes tales firmitates disfactas habeant; quia vicini et
circummanentes exinde multas depredationes et impedimenta sustinent."
_Capitularia Regum Francorum_, Boretius, ii., 328.

[174] These instances are as follows:--868, A certain Acfrid shut
himself up in a _casa firmissima_ in the _villa_ of Bellus Pauliacus
on the Loire, and it was burnt over his head (_Annales Bertinianorum_,
pp. Migne, 125, 1237); 878, The sons of Goisfrid attack the _castellum_
and lands of the son of Odo (_ibid._, p. 1286); 879, Louis the Germanic
besieges some men of Hugh, son of Lothaire, _in quodam castello
juxta Viridunum_: he takes and destroys the castellum (_Annals of
Fulda_, Pertz, i., 393); 906, Gerard and Matfrid fortify themselves
in a certain _castrum_, in a private war (_Regino_, Pertz, i., 611).
Sismondi states that the great nobles wrested from Louis-le-Bégue
(877-879) the right of building private castles. So far, we have been
unable to find any original authority for this statement.

[175] See Guizot, _Histoire de la Civilisation_, iii., 309. "On voit
les _villæ_ s'entourer peu à peu de fossés, de remparts de terre, de
quelques apparences de fortifications."

[176] We hear of monasteries being fortified in this way; in 869
Charles the Bald drew a bank of wood and stone round the monastery
of St Denis; "castellum in gyro ipsius monasterii ex ligno et lapide
conficere coepit." _Ann. Bertinian_, Migne, pp. 125, 1244. In 889 the
Bishop of Nantes made a _castrum_ of his church by enclosing it with a
wall, and this wall appears to have had a tower. _Chron. Namnetense_,
p. 45, in _Lobineau's Bretagne_, vol. ii. In 924 Archbishop Hervey made
a _castellum_ of the monastery of St Remi by enclosing it with a wall.
_Flodoard_, p. 294 (Migne). But the fortification of monasteries was a
very different thing from the fortification of private castles.

[177] In 951 Duke Conrad, being angry with certain men of Lorraine,
threw down the _towers_ of some of them; these may have been the keeps
of private castles. Flodoard, _Annales_, p. 477.

[178] _Presidium_ is one of those vague words which chroniclers love to
use; it means a defence of any kind, and may be a town, a castle, or a
garrison. The town in which this turris stood appears by the context to
have been Chateau Thierry. _Cf._ Flodoard, _Annales_, pp. 924, with 933.

[179] "Castrum muro factum circa eam [ecclesiam]." _Chron. Namnetense_,
p. 45. "Precepit [Alanus] eis terrarium magnum in circuitu Ecclesiæ
facere, sicut murus prioris castri steterat, quo facto turrem
principalem _reficiens_, in ea domum suam constitit." _Ibid._

[180] Flodoard, _Annales_, pp. 931 and 949. This tower was heightened
by Charles, the last of the Carlovingians, and furnished with a ditch
and bank, in 988.

[181] It is often supposed that these towers were derived from the
_Pretoria_, or general's quarters in the Roman _castra_. It is far more
probable that they were derived from mural towers. The Pretorium was
not originally fortified, and it was placed in the centre of the Roman
camp. But one great object of the feudal keep was to have communication
with the open country. The keep of Laon was certainly on the line of
the walls, as Bishop Ascelin escaped from it down a rope in 989, and
got away on a horse which was waiting for him. Palgrave, _England and
Normandy_, ii., 880.

[182] The word _motte_ or _mota_ does not occur in any contemporary
chronicle, as far as is known to the writer, before the 12th century;
nor is the word _dangio_ to be found in any writer earlier than
Ordericus. But the _thing_ certainly existed earlier.

[183] [Fulk and his son Geoffrey] "in occidentali parte montis castellum
determinaverunt.... Aggerem quoque in prospectu monasterii cum turre
lignea erexerunt." _Chron. St Florentii_, in Lobineau's _Bretagne_,
ii., 87. Some remains of this motte are still visible. De Salies,
_Foulques Nerra_, p. 263.

[184] "Elegantissimus in rebus bellicis" is the quaint language of the
Angevin chronicler, 176.

[185] See De Salies, _Histoire de Foulques Nerra_, which indirectly
throws considerable light on the archæological question.

[186] Salies, _Histoire de Foulques Nerra_, p. 170. M. Enlart, in his
_Manuel d'Archæologie Française_, ii., 495, has been misled about this
castle by the _Chronicon Andegavense_, which says: "Odo.... Fulconem
expugnare speravit, et totis nisibus adorsus est. Annoque presenti
(1025) Montis Budelli castellum, quod circiter annos decem retro abhinc
contra civitatem Turonicam firmaverat Fulco, obsedit, et turrim ligneam
miræ altitudinis super domgionem ipsius castri erexit." Bouquet, x.,
176. M. Enlart takes this to be the first recorded instance of a
motte. But the passage is evidently corrupt, as the other accounts of
this affair show that Count Odo's wooden tower was a siege engine,
employed to attack Fulk's castle, and afterwards burnt by the besieged.
See the _Gesta Ambasiens. Dom._, _ibid._, p. 257, and the _Chron. St
Florentii_. Probably we should read _contra_ domgionem instead of
_super_. The _Chronicon Andegavense_ was written in the reign of Henry
II.

[187] When Fulk invaded Bretagne in or about 992, he collected an army
"tam de suis quam conductitiis." _Richerius_, edition Guadet. The
editor remarks that this is perhaps the first example of the use of
mercenaries since the time of the Romans (ii., 266). Spannagel, citing
Peter Damian, says that mercenaries were already common at the end of
the 10th century. _Zur Geschichte des Deutschen Heerwesens_, pp. 72, 73.

[188] This was always the custom in mediæval castles. See Cohausen,
_Befestigungen der Vorzeit_, p. 282.

[189] "Qui vivens turres altas construxit et ædes, Unam Carnotum, sed
apud Dunense reatum." _Chron. St Florentii._

[190] _Chron. Namnetense_, Lobineau, ii., 47.

[191] _Gesta Ambasiensium Dominorum_, in _Spicilegium_, p. 273.

[192] _Guide Joanne_, p. 234.

[193] The furthest point of the headland on which the castle is
placed is a small circular court, with a fosse on all sides but the
precipices. From personal visitation.

[194] _Dunio_ is subsequently explained by Lambert as _motte_: "Motam
altissimam sive dunionem eminentem in munitionis signum firmavit."
_Lamberti Ardensis_, p. 613. It is the same word as the Saxon _dun_,
a hill (preserved in our South Downs), and has no connection with the
Irish and Gaelic _dun_, which is cognate with the German _zaun_, a
hedge, A.-S. _tun_, and means a hedged or fortified place. The form
_dange_ appears in Northern France, and this seems to be the origin of
the word _domgio_ or _dangio_ which we find in the chroniclers, the
modern form of which is _donjon_. If we accept this etymology, we must
believe that the word _dunio_ or _domgio_ was originally applied to
the hill, and not to the tower on the hill, to which it was afterwards
transferred. It is against this view that Ordericus, writing some
fifty years before Lambert, uses the form _dangio_ in the sense of a
tower. Professor Skeat and the _New English Dictionary_ derive the word
_donjon_ or _dungeon_ from Low Lat. _domnionem_, acc. of _domnio_, thus
connecting it with _dominus_, as the seignorial residence.

[195] Ducange conjectured that the motte-castle took its origin in
Flanders, but it was probably the passage cited above from Lambert
which led him to this conclusion. See art. "Mota" in Ducange's
_Glossarium_.

[196] Steenstrup, _Normannerne_, i., 297.

[197] _Ibid._, i., 301.

[198] _England and Normandy_, ii., 535.

[199] "Muros et propugnacula civitatum refecit et augmentavit." _Dudo_,
p. 85 (Duchesne's edition).

[200] "Henricus rex circa turrem Rothomagi, _quam ædificavit primus
Richardus dux Normannorum in palatium sibi_, murum altum et latum cum
propugnaculis ædificat." _Robert of Toringy_, R.S., p. 106.

[201] _Ordericus_, ii., 15, 17, 46 (edition Prévost).

[202] _William of Jumièges_, anno 1035. Mr Freeman remarks that the
language of William would lead us to suppose that the practice of
castle-building was new.

[203] There are some facts which render it probable that the earliest
castles built in Normandy were without mottes, and were simple
enclosures like those we have described already. Thus the castle of
the great family of Montgomeri is an enclosure of this simple kind.
Domfront, built by William Talvas in Duke Robert's time, has no motte.
On the other hand, Ivry, built by the Countess Albereda in Duke Richard
I.'s days, "on the top of a hill overlooking the town" (William of
Jumièges), may possibly have been a motte; and there is a motte at
Norrei, which we have just mentioned as an early Norman castle.

[204] _Manuel d'Archæologie Française_, p. 457.

[205] This want will be supplied, as regards England, by the completion
of the _Victoria County Histories_, and as regards France, by the
_Societé Préhistorique_, which is now undertaking a catalogue of all
the earthworks of France. The late M. Mortillet, in an article in the
_Revue Mensuelle de l'École d'Anthropologie_, viii., 1895, published
two lists, one of actual mottes in France, the other of place-names in
which the word motte is incorporated. Unfortunately the first list is
extremely defective, and the second, as it only relates to the name,
is not a safe guide to the proportional numbers of the thing. All that
the lists prove is that mottes are to be found in all parts of France,
and that place-names into which the word _motte_ enters seem to be more
abundant in Central France than anywhere else. It is possible that a
careful examination of local chroniclers may lead to the discovery
of some earlier motte-builder than Thibault-le-Tricheur. We should
probably know more about Thibault's castles were it not that the Pays
Chartrain, as Palgrave says, is almost destitute of chroniclers.

[206] Cited at length by De Caumont, _Bulletin Monumental_, ix., 246.
Von Hochfelden considered that the origin of feudal fortresses in
Germany hardly goes back to the 10th century; only great dukes and
counts then thought of fortifying their manors; those of the small
nobility date at earliest from the end of the 12th century.

[207] _Die Befestigungen der Vorzeit_, p. 28.

[208] _Entwickelung des Kriegswesens_, iii., 370.

[209] _Antiquitates Italicæ_, ii., 504. He says they are many times
mentioned both in charters and chronicles in Italy.

[210] We hear of Robert Guiscard building a wooden castle on a hill at
Rocca di St Martino in 1047. Amari, _Storia dei Musulmani di Sicilia_,
i., 43. Several place-names in Italy and Sicily are compounded with
_motta_, as the Motta Sant'Anastasia in Sicily. See Amari, _ibid._, p.
220.

[211] Especially Montfort and Blanchegarde. But there is a wide field
for further research both in Palestine and Sicily.

[212] "Bei den Sclaven haben die Chateaux-à-motte keinen Eingang
gefunden, weil ihnen das Lehnswesen fremd geblieben ist." iii., 338.

[213] Professor Montelius informed the writer that they are quite
unknown in Norway or Sweden; and Dr Christison obtained an assurance to
the same effect from Herr Hildebrand.

[214] "These are small well-defended places, the stronghold of the
individual, built for a great man and his followers, and answering to
mediæval conditions, to a more or less developed feudal system." _Vor
Oldtid_, p. 642.

[215] I am informed by a skilled engineer that even in the wet climate
of England it would take about ten years for the soil to settle
sufficiently to bear a stone building.

[216] Köhler says: "By far the greater part of the castles of the
Teutonic knights in Prussia, until the middle of the fourteenth
century, were of wood and earth." _Die Entwickelung des Kriegswesen_,
iii., 376.

[217] _Cal. of Patent Rolls_, 1232-1247, p. 340. Mandate to provost of
Oléron to let Frank De Brene have tools to make a new motte in the isle
of Rhé. Later the masters and crews of the king's galleys are ordered
to help in building the motte and the wooden castle. P. 343.

[218] _Antiquitates Italicæ_, ii., 504. Can Grande's motte at Padua.
Anno 1320. "Dominus Alternerius [podesta of Padua] ... cum maxima
quantitate peditum et balistariorum Civitatis Paduæ, iverunt die
predicto summo mane per viam Pontis Corvi versus quamdam motam magnam,
quam faciebat facere Dominus Canis, cum multis fossis et tajatis ad
claudendum Paduanos, ne exirent per illam partem, et volendo ibidem
super illam motam ædificare castrum. Tunc prædictus Potestas cum aliis
nominatis splanare incoeperunt, et difecerunt dictam motam cum tajatis
et fossa magna."

We may remark here that as early as the 17th century the learned
Muratori protested against the equation of _mota_ and _fossatum_,
and laughed at Spelman for making this translation of _mota_ in his
_Glossary_. _Antiquitates Italicæ_, ii., 504.

[219] Cited by Westropp, _Journal of R.S.A., Ireland_, 1904.

[220] Vicars' _Parliamentary Chronicle_, cited by Hunter, _South
Yorks_, ii., 235.

[221] "Camps on the Malvern Hills," _Journ. Anthrop. Inst._, x., 319.

[222] M. de Salies has traced in detail the connection between Fulk
Nerra's castles and the Roman roads of Anjou and Touraine.

[223] See some excellent remarks on this subject in Mr W. St John
Hope's paper on "English Fortresses" in _Arch. Journ._, lx., 72-90.

[224] Only a very small number of mottes have as yet been excavated.
Wells were found at Almondbury, Berkeley, Berkhampstead, Carisbrook,
Conisborough, Kenilworth, Northallerton, Norwich, Pontefract, Oxford,
Tunbridge, Worcester, and York. At Caus, there is a well in the ditch
between the motte and the bailey. Frequently there is a second well in
the bailey.

[225] The writer at one time thought that the ruins at the east end of
the castle of Pontefract concealed a second motte, but wishes now to
recant this opinion. _Eng. Hist. Review_, xix., 419.

[226] Thus Henry I. erected a siege castle to watch Bridgenorth
(probably Pampudding Hill), and then went off to besiege another
castle. Mr Orpen kindly informs me that the camp from which Philip
Augustus besieged Château Gaillard contains a motte. Outside Pickering,
Corfe, and Exeter there are earthworks which have probably been siege
castles.

[227] Henry II. built a castle and very fine borough (burgum pergrande)
at Beauvoir in Maine. _Robert of Torigny_, R.S., p. 243. Minute
regulations concerning the founding of the borough of Overton are given
in _Close Rolls_, Edward I. (1288-1296), p. 285.

[228] See Round, _Studies in Domesday_, pp. 125, 126.

[229] Neckham, "De Utensilibus," in Wright's _Volume of Vocabularies_,
pp. 103, 104. Unfortunately this work of Neckham's was not written to
explain the construction of motte castles, but to furnish his pupils
with the Latin names of familiar things; a good deal of it is very
obscure now.

[230] See frontispiece.

[231] _Acta Sanctorum_, 27th January, Bolland, iii., 414. This
biography was written only nine months after Bishop John's death, by an
intimate friend, John de Collemedio.

[232] Guisnes is now in Picardy, but in the 12th century it was in
Flanders, which was a fief of the Empire.

[233] This description is from the _Historia Ardensium_ of Walter de
Clusa, which is interpolated in the work of Lambert, Bouquet, pp. 13,
624.

[234] Yet in some of the later keeps, such as Conisburgh, where we find
only one window to a storey, the room must have been undivided.

[235] See Wright, _History of Domestic Manners_, p. 26.

[236] According to Littré, the original derivation of the word _motte_
is unknown. I have not found any instance of the word _mota_ in
chronicles earlier than the 12th century, but the reason appears to be
that _mota_ or motte was a folk's word, and appeared undignified to an
ambitious writer. Thus the author of the _Gesta Consulum Andegavensium_
says that Geoffrey Martel, Count of Anjou, gave to a certain Fulcoius
the fortified house which is still called by the vulgar Mota Fulcoii.
D'Achery, _Spicilegium_, p. 257.

[237] See Appendix G.

[238] See Appendix H.

[239] _Peel, its Meaning and Derivation_, by George Neilson.

[240] See Appendix I. Cohausen has some useful remarks on the use of
hedges in fortification. _Befestigungen der Vorzeit_, pp. 8-13. A
quickset hedge had the advantage of resisting fire. The word _sepes_,
which properly means a hedge, is often applied to the palitium.

[241] This list or _catalogue raisonné_ was originally published in the
_English Historical Review_ for 1904 (vol. xix.). It is now reproduced
with such corrections as were necessary, and with the addition of five
more castles, as well as of details about thirty-four castles for which
there was not space in the _Review_. The Welsh castles are omitted from
this list, as they will be given in a separate chapter.

[242] The list is brought up to fifty by interpreting the _regis domus_
of Winchester to be Winchester castle; the reasons for this will be
given later. The number would be increased to fifty-two if we counted
Ferle and Bourne in Sussex as castles, as Mr Freeman does in his
_Norman Conquest_, v., 808. But the language of Domesday seems only to
mean that the lands of these manors were held of Hastings castle by the
service of castle-guard. See D. B., i., pp. 21 and 206.

[243] The total number would be eighty-six if Burton and Aldreth were
included. Burton castle is mentioned in Domesday, but there is no
further trace of its existence. The castle of Alrehede or Aldreth in
the island of Ely is stated by the _Liber Eliensis_ to have been built
by the Conqueror, but no remains of any kind appear to exist now. Both
these castles are therefore omitted from the list.

[244] Exact numbers cannot be given, because in some cases the bounds
of the ancient borough are doubtful, as at Quatford.

[245] At Winchester and Exeter. For Winchester, see Milner, _History of
Winchester_, ii., 194; for Exeter, Shorrt's _Sylva Antiqua Iscana_, p.
7.

[246] Colchester is the only exception to this rule, as the castle
there is in the middle of the town; but even this is only an apparent
exception, as the second bailey extended to the town wall on the north,
and had been royal demesne land even before the Conquest. See Round's
_Colchester Castle_, ch. vii.

[247] These five are Berkeley, Berkhampstead, Bourn, Pontefract,
Rayleigh.

[248] I am indebted for these measurements to Mr D. H. Montgomerie.

[249] Notification in Round's _Calendar of Documents preserved in
France_, p. 367. Mr Round dates the Notification 1087-1100.

[250] Description furnished by Mr D. H. Montgomerie, F.S.A.

[251] "Castrum Harundel T. R. E. reddebat de quodam molino 40 solidos,
et de 3 conviviis 20 solidos, et de uno pasticio 20 solidos. Modo inter
burgum et portum aquæ et consuetudinem navium reddit 12 libras, et
tamen valet 13. De his habet S. Nicolaus 24 solidos. Ibi una piscaria
de 5 solidos et unum molinum reddens 10 modia frumenti, et 10 modia
grossæ annonæ. Insuper 4 modia. Hoc appreciatum est 12 libras. Robertus
filius Tetbaldi habet 2 hagas de 2 solidis, et de hominibus extraniis
habet suum theloneum." Several other _hagæ_ and _burgenses_ are then
enumerated. (D. B., i., 23a, 1.)

[252] See Mr Round's remarks on the words in his _Geoffrey de
Mandeville_, Appendix O. The above was written before the appearance
of Mr Round's paper on "The Castles of the Conquest" (_Archæologia_,
lviii.), in which he rejects the idea that _castrum Harundel_ means the
castle.

[253] See _ante_, p. 28.

[254] Florence of Worcester mentions the castle of Arundel as belonging
to Roger de Montgomeri in 1088.

[255] See Appendix R.

[256] The expenses entered in the _Pipe Rolls_ (1170-1187) are for the
works of the castle, the chamber and wall of the castle, the _houses_
of the castle (an expression which generally refers to the keep), and
for flooring the tower (turris) and making a garden. _Turris_ is the
usual word for a keep, and is never applied to a mere mural tower.

[257] This gateway is masked by a work of the 13th century, which
serves as a sort of barbican.

[258] In operibus castelli de Arundel 22_l._ 7_s._ 8_d._ Et debet
55_l._ 18_s._ 6_d._ _Pipe Roll_, 31, Henry I., p. 42.

[259] D. B., i., 23a, 1.

[260] _Testa de Nevill_, i., iii., 236, cited by C. Bates, in a very
valuable paper on Bamborough Castle, in _Archæologia Æliana_, vol.
xiv., "Border Holds." Mr Bates gives other evidence to the same
effect. The early existence of the castle is also proved by the fact
that Gospatric, whom William had made Earl of Northumberland, after
his raid on Cumberland in 1070, brought his booty to the _firmissimam
munitionem_ of Bamborough. Symeon of Durham, 1070.

[261] _Vita S. Oswaldi_, ch. xlviii., in Rolls edition of Symeon.

[262] This was the opinion of the late Mr Cadwalader Bates, who thought
that the smallness of the sums entered for Bamborough in Henry II.'s
reign might be accounted for by the labour and materials having been
furnished by the crown tenants. _Border Strongholds_, p. 236.

[263] Bamborough rock has every appearance of having been once an
island. As late as 1547 the tide came right up to the rock on the east
side; the sea is now separated from the castle by extensive sandhills.

[264] _M. A._, v., 197.

[265] _Domesday_ mentions the destruction of twenty-three houses at
Barnstaple, which may have been due partly or wholly to the building of
the castle. I., 100.

[266] From a lecture by Mr J. R. Chanter.

[267] The _Fundatio_ of Belvoir priory says that Robert founded the
church of St Mary, _juxta castellum suum_, _M. A._, iii., 288. As
Robert's coffin was actually found in the Priory in 1726, with an
inscription calling him Robert de Todnei _le Fundeur_, the statement is
probably more trustworthy than documents of this class generally are.

[268] Nicholls, _History of Leicester_, i., 110.

[269] D. B., i., 233b.

[270] "In Ness sunt 5 hidæ pertinentes ad Berchelai, quas comes
Willielmus misit extra ad faciendum unum castellulum." D. B., i., 163a,
2.

[271] "Castella per loca firmari præcepit." _Flor. Wig._, 1067. See
Freeman, _N. C._, iv., 72. Domesday tells us that FitzOsbern built
Ness, Clifford, Chepstow, and Wigmore, and rebuilt Ewias.

[272] Robert Fitzhardinge, in his charter to St Austin's Abbey at
Bristol, says that King Henry [II.] gave him the manor of Berchall, and
all Bercheleiernesse. _Mon. Ang._, vi., 365.

[273] It is not necessary to discuss the authenticity of the story
preserved by Walter Map; it is enough that Gytha, the wife of Godwin,
held in horror the means by which her husband got possession of
Berkeley Nunnery. D. B., i., 164.

[274] _Mediæval Military Architecture_, i., 236.

[275] The gift of the manor was made before Henry became king, and was
confirmed by charter on the death of Stephen in 1154. Fitzhardinge
was an Englishman, son of an alderman of Bristol, who had greatly
helped Henry in his wars against Stephen. See Fosbroke's _History of
Gloucester_.

[276] He held Berkeley under the crown at the time of the Survey. D.
B., i., 163a.

[277] From information received from Mr Duncan Montgomerie.

[278] Fosbroke's _History of Gloucester_ attributes this bailey to
Maurice, son of Robert Fitzhardinge. One of the most interesting
features in this highly interesting castle is the wooden pentice
leading from the main stairway of the keep to the chamber called Edward
II.'s. Though a late addition, it is a good instance of the way in
which masonry was eked out by timber in mediæval times.

[279] Clark, _M. M. A._, i., 229.

[280] D. B., i., 163.

[281] _Victoria County History of Herts_, from which the description of
these earthworks is entirely taken.

[282] _Mon. Ang._, vii., 1090.

[283] They were excavated by Mr Montgomerie in 1905, and no trace of
masonry was found.

[284] Roger of Wendover, 1216.

[285] D. B., i., 163.

[286] The charter, which is in both Anglo-Saxon and Latin, is given in
Dugdale's _History of St Paul's_, 304.

[287] See _Freeman_, ii., 356; and D. B., i., 134a.

[288] From report by Mr D. H. Montgomerie.

[289] _Waytemore_ has sometimes been identified with the puzzling
Wiggingamere, but in defiance of phonology.

[290] D. B., i., 351b.

[291] _M. A._, vi., 86.

[292] _Itin._, i., 27.

[293] _Associated Archæological Societies_, VI., ix.

[294] Report by Mr D. H. Montgomerie.

[295] Ipse Willielmus tenet Wasingetune. Guerd Comes tenuit T. R. E.
Tunc se defendebat pro 59 hidis. Modo non dat geldum. In una ex his
hidis sedet castellum Brembre. D. B., i., 28a, 1.

[296] We often find that the architecture of the nearest church throws
light on the date of the castle. A Norman seldom built or restored his
castle without doing something for the church at the same time.

[297] See Ordericus, ii., 178.

[298] The _Chronica de Fundatoribus of Tewkesbury Abbey_ seems to be
the origin of the tradition that Earl Robert was the builder of Bristol
Castle. There can be no doubt that his work was in stone, as the same
authority states that he gave every tenth stone to the Chapel of Our
Lady in St James' Priory. _M. A._, ii., 120. According to Leland, the
keep was built of Caen stone. _Itin._, vii., 90. Robert of Gloucester
calls it the flower of all the towers in England.

[299] We have no historical account of the Norman conquest of Bristol,
and the city is only mentioned in the most cursory manner in D. B.

[300] Seyer (_Memoirs of Bristol_, i.) was convinced that the plan
published by Barrett, and attributed to the monk Rowlie, was a forgery;
his own plan, as he candidly admits, was largely drawn from imagination.

[301] Castellum plurimo aggere exaltatum. _Gesta Stephani_, 37.

[302] Seyer, i., 391, and ii., 82.

[303] Quoted by Seyer, ii., 301, from _Prynne's Catal._, p. 11.

[304] Calculated from the measurements given by William of Worcester.
_Itin._, p. 260. William probably alludes to the motte when he speaks
of the "mayng round" of the castle.

[305] _Benedict of Peterborough_, i., 92.

[306] _Hist. of Bristol_, i., 373.

[307] _Ibid._, vol. ii.

[308] _De Gestis Herewardi Saxonis_, Wright's edition. See Freeman, N.
C., iv., 804.

[309] _Beauties of England and Wales_, Buckingham, p. 282.

[310] Camden's _Britannia_, i., 315.

[311] D. B., i., 143.

[312] "Willielmus de Scohies tenet 8 carucatas terræ in castellaria de
Carliun, et Turstinus tenet de eo. Ibi habet in dominio unam carucam,
et tres Walenses lege Walensi viventes, cum 3 carucis, et 2 bordarios
cum dimidio carucæ, et reddunt 4 sextares mellis. Ibi 2 servi et una
ancilla. Hæc terra wasta erat T. R. E., et quando Willelmus recepit.
Modo valet 40 solidos." D. B., i., 185b, 1.

[313] The _Gwentian Chronicle_, Cambrian Archæological Association,
A.D. 962, 967. It is not absolutely impossible that these passages
refer to Chester. Caerleon appears to have been seized by the Welsh
very soon after the death of William I.

[314] _Itin. Camb._, p. 55.

[315] Loftus Brock, in _Journ. Brit. Arch. Ass._, xlix. J. E. Lee, in
_Arch. Camb._, iv., 73.

[316] D. B., i., 185b.

[317] [Rex] "in reversione sua Lincolniæ, Huntendonæ et Grontebrugæ
castra locavit." _Ord. Vit._, p. 189.

[318] D. B., i., 189.

[319] A similar plan was made independently by the late Professor
Babington. Some traces of the original earthwork of the city are still
to be seen. See Mr Hope's paper on _The Norman Origin of Cambridge
Castle_, Cambridge Antiquarian Soc., vol. xi.; and Babington's _Ancient
Cambridgeshire_, in the same society's _Octavo Publications_, No. iii.,
1853.

[320] W. H. St John Hope, as above, p. 342.

[321] "Archiepiscopus habet ex eis [burgensibus] 7 et abbas S.
Augustini 14 pro excambio castelli." D. B., i. a, 2.

[322] "Et undecim sunt perditi infra fossatum castelli"; cited by
Larking, _Domesday of Kent_, App. xxiv. Domesday says, "sunt vastatæ
xi. in fossa _civitatis_." There can be no doubt that the Chartulary
gives the correct account.

[323] The hill is called the Dungan, Dangon, or Dungeon Hill in many
old local deeds. See "Canterbury in Olden Times,"_Arch. Journ._, 1856.
Stukeley and Grose both call it the Dungeon Hill.

[324] See Appendix N.

[325] Somner's _Antiquities of Canterbury_, p. 144. Published in 1640.

[326] _Antiquities of Canterbury_, p. 75.

[327] Mr Clark thought there was another motte in the earthworks
outside the walls, though he expresses himself doubtfully: "I rather
think they [the mounds outside the city ditch] or one of them, looked
rather like a moated mound, but I could not feel sure of it." _Arch.
Cantiana_, xv., 344. Gostling (_A Walk about Canterbury_, 1825) says
there were _two_, which is perhaps explained by a passage in Brayley's
_Kent_ (1808), in which he describes the external fortification as "a
lesser mount, now divided into two parts, with a ditch and embankment."
P. 893. Stukeley's description (circa 1700) is as follows: "Within
the walls is a very high mount, called Dungeon Hill; a ditch and high
bank enclose the area before it; it seems to have been part of the old
castle. Opposite to it without the walls is a hill, seeming to have
been raised by the Danes when they besieged the city. The top of the
Dungeon Hill is equal to the top of the castle." _Itin. Curiosum_,
i., 122. It is of course not impossible that there may have been two
mottes to this castle, as at Lewes and Lincoln, but such instances
are rare, and it seems more likely that a portion of the bailey bank
which happened to be in better preservation and consequently higher
was mistaken for another mount. Mr Clark committed this very error at
Tadcaster, and the other writers we have quoted were quite untrained
as observers of earthen castles. At any rate there can be no doubt
that the Dane John is the original chief citadel of this castle,
as the statements of Somner, Stukeley, and we may add, Leland, are
explicit. The most ancient maps of Canterbury, Hoefnagel's (1570),
Smith's (_Description of England_, 1588), and Grose's (1785), all show
the Dungeon Hill within the walls, but take no notice of the outwork
outside.

[328] _Archæologia Cantiana_, xxxiii., 152.

[329] _Ibid._, xxi.

[330] _Close Rolls_, i., 234b, ii., 7b, 89.

[331] Now, to the disgrace of the city of Canterbury, converted into
gasworks.

[332] For instance, at Middleham, Rochester, Rhuddlan, and Morpeth.

[333] _Beauties of England and Wales, Kent_, p. 893.

[334] The passages from the _Pipe Roll_ bearing on this subject (which
have not been noticed by any previous historian of Canterbury) are as
follows:--

 1166-7. In operatione civitatis Cantuar. claudendæ          £5   19   6

    "    Ad claudendam civitatem Cantuar.                    20    0   0

 1167-8. Pro claudenda civitate Cantuar.                      5    1   1

 1168-9. In terris datis Adelizæ filie Simonis 15 solidos
           de tribus annis pro escambio terræ suæ quæ est
           in Castello de Cantuar.                            0   15   0

 1172-3. In operatione turris ejusdem civitatis              10    0   0

    "    In operatione predicte turris                       53    6   8

    "    Summa denariorum quos vicecomes misit in
           operatione turris                                 73    1   4

 1173-4. In operatione turris et Castelli Chant.             24    6   0

    "    In operatione turris Cantuar.                        5   11   7

 1174-5. Et in warnisione ejusdem turris                      5    8   0

The latter extract, which refers to the provisioning of the keep, seems
to show that it was then finished. The sums put down to the castle,
amounting to about £4000 of our money, are not sufficient to defray the
cost of so fine a keep. But the entries in the _Pipe Rolls_ relate only
to the Sheriff's accounts, and it is probable that the cost of the keep
was largely paid out of the revenues of the archbishopric, which Henry
seized into his own hands during the Becket quarrel.

[335] The portion of the wall of Canterbury, which rests on an earthen
bank, extends from Northgate to the Castle, and is roughly semicircular
in plan. In the middle of it was St George's Gate, which was anciently
called _Newingate_ (Gostling, p. 53) and may possibly have been Henry
II.'s new gate. The part enclosing the Dungeon Hill is angular, and
appeared to Mr Clark, as well as to Somner and Hasted, to have been
brought out at this angle in order to enclose the hill.

[336] _Arch. Journ._, 1856.

[337] D. B., i., 2a, 1.

[338] "Isdem rex tenet Alwinestone. Donnus tenuit. Tunc pro duabus
hidis et dimidia. Modo pro duabus hidis, quia castellum sedet in una
virgata." D. B., i., 2a, 1.

[339] See below, under Windsor.

[340] "In hac [insula] castellum habebat ornatissimum lapidum ædificio
constructum, validissimo munimine firmatum." _Gesta Stephani_, R. S.,
p. 28.

[341] Stone's _Official Guide to the Castle of Carisbrooke_, p. 39.

[342] Mr W. H. Stevenson, in his edition of _Asser_, pp. 173, 174,
shows that the name Carisbrooke cannot possibly be derived from
Wihtgares-burh, as has been sometimes supposed, as the older forms
prove it to have come from _brook_, not _burh_. The lines of the
present castle banks, if produced, would not correspond with those of
the Tilt-yard, which is proof that the Norman castle was not formed by
cutting an older fortification in two.

[343] Bower's _Scotochronicon_, v., xlii. Cited by Mr Neilson, _Notes
and Queries_, viii., 321. See also Palgrave, _Documents and Records_,
i., 103.

[344] _Cal. of Close Rolls_, Edward II., iii., 161.

[345] _Mon. Ang._, v., 12. "Castelli nostri de Acra."

[346] As at Burton, Mexborough, Lilbourne, and Castle Colwyn.

[347] Harrod's _Gleanings among the Castles and Convents of Norfolk_.
See also _Arch. Journ._, xlvi., 441.

[348] D. B., ii., 160b.

[349] "Castellum de Estrighoiel fecit Willelmus comes, et ejus tempore
reddebat 40 solidos, tantum de navibus in silvam euntibus." D. B.,
i., 162. Tanner has shown that while Chepstow was an alien priory
of Cormeille, in Normandy, it is never spoken of by that name in
the charters of Cormeille, but is always called Strigulia. _Notitia
Monastica_, Monmouthshire. See also Marsh's _Annals of Chepstow Castle_.

[350] I must confess that in spite of very strong opposing opinions, I
see no reason why this building should not be classed as a keep. It is
of course a gross error to call Martin's Tower the keep; it is only a
mural tower.

[351] D. B., 162, 1a.

[352] "Cestriæ munitionem condidit." P. 199 (Prévost's edition).

[353] _Chester Historical and Archæological Society_, v., 239.

[354] _Pipe Rolls_, ii., 7. Ranulph, Earl of Chester, died in 1153, and
the castle would then escheat into the king's hands.

[355] This work seems to have been completed in the reign of Edward
II., who spent £253 on the houses, towers, walls, and gates. _Cal. of
Close Rolls_, Edward II., ii., 294.

[356] _Close Rolls_, 35, Henry III., cited by Ormerod, _History of
Cheshire_, i., 358.

[357] See Mr Cox's paper, as above, and Shrubsole, _Chester Hist. and
Arch. Soc._, v., 175, and iii., New Series, p. 71.

[358] _Benedict of Peterborough_, i., 135, R. S.

[359] D. B., i., 262b.

[360] "Willelmus comes fecit illud [castellum] in wasta terra quam
tenebat Bruning T. R. E." D. B., i., 183a, 2.

[361] "Ancient Charters," _Pipe Roll Society_, vol. x., charter xiii.,
and Mr Round's note, p. 25.

[362] It is extraordinary that Mr Clark, in his description of this
castle, does not mention the motte, except by saying that the outer
ward is 60 or 70 feet lower than the inner. _M. M. A._, i., 395.

[363] This passage occurs in a sort of appendix to Domesday Book, which
is said to be in a later hand, of the 12th century. (Skaife, _Yorks.
Arch. Journ._, Part lv., p. 299.) It cannot, however, be very late in
the 12th century, as it speaks of Roger's holdings in Craven in the
present tense.

[364] See Farrer's _Lancashire Pipe Rolls_, p. 385. The castle is not
actually mentioned, but "le Baille" (the bailey) is spoken of. Mr
Farrer also prints an abstract of a charter of Henry I. (1102): "per
quam concessit eidem Roberto [de Laci] Boelandam [Bowland] quam tenuit
de Rogero Comite Pictavensi, ut extunc eam de eodem rege teneat." P.
382.

[365] In an inquisition of Henry de Laci (+ 1311) it is said that
"castelli mote et fossæ valent nihil." (Whitaker's _History of
Whalley_, p. 280.) This is probably an instance of the word _motte_
being applied to a natural rock which served that purpose. See another
instance under Nottingham, _post_, p. 176.

[366] Dugdale's _Baronage_, i., p. 99. Dugdale's authority appears to
have been the "Historia Laceiorum," a very untrustworthy document,
but which may have preserved a genuine tradition in this instance.
The loopholes in the basement of the keep, with the large recesses,
appear to have been intended for crossbows, and the crossbow was not
reintroduced into England till the reign of Richard I.

[367] _Victoria History of Lancashire_, ii., 523.

[368] See Farrer, _Lancashire Pipe Rolls_, i., 260.

[369] Printed by Mr Round in _Essex Arch. Society's Transactions_,
vii., Part ii. The charter is dated 1101.

[370] See Maitland, _Domesday Book and Beyond_, p. 22.

[371] _History of Colchester Castle_, p. 141.

[372] It has been much debated whether these tiles are Roman or Norman;
the conclusion seems to be that they are mixed. See Round's _History of
Colchester_, p. 78.

[373] The single _Pipe Roll_ of Henry I. shows that he spent £33, 15s.
on repairs of the castle and borough in 1130.

[374] In operatione unius Rogi (a kiln), £13, 18s. In reparatione muri
castelli, £16, 3s. 2d. The projection of the buttresses (averaging 1
ft. 3 ins.) is about the same as that found in castles of Henry I. or
Henry II.'s time.

[375] Ad faciendum Ballium circa castellum, £50. _Pipe Rolls_, xix.,
13. This is followed by another entry of £18, 13s. 7d. "in operatione
castelli," which may refer to the same work.

[376] Round's _History of Colchester_.

[377] _Close Rolls_, i., 389. Mandamus to the bishop of London to
choose two lawful and discreet men of Colchester, "et per visum
eorum erigi faceatis palicium castri nostri Colecestrie, quod nuper
prostratum fuit per tempestatem."

[378] Round's _History of Colchester_, pp. 135, 136.

[379] Tota civitas ex omnibus debitis reddebat T. R. E., £15, 5s. 4d.,
in unoquoque anno. Modo reddit £160. D. B., ii., 107.

[380] Eyton, _Key to Domesday_, p. 43. This passage was kindly pointed
out to me by Dr Round. The castle is not mentioned in Domesday under
Wareham, but under Kingston. "De manerio Chingestone habet rex unam
hidam, in qua fecit castellum Warham, et pro ea dedit S. Mariæ [of
Shaftesbury] ecclesiam de Gelingeham cum appendiciis suis." D. B., i.,
78b, 2.

[381] "Advocatio ecclesie de Gillingeham data fuit abbati [_sic_] de
S. Edwardo in escambium pro terra ubi castellum de Corf positum est."
_Testa de Nevill_, 164b.

[382] It is by no means certain that Corfe was the scene of Edward's
murder, as we learn from a charter of Cnut (_Mon. Ang._, iii., 55) that
there was a Corfe Geat not far from Portisham, probably the place now
called Coryates.

[383] Called by Asser a _castellum_; but it has already been pointed
out that _castellum_ in early writers means a walled town and not a
castle. (See p. 25.) Wareham is a town fortified by an earthen vallum
and ditch, and is one of the boroughs of the _Burghal Hidage_. (See Ch.
II, p. 28.) A Norman castle was built there after the Conquest, and
its motte still remains. D. B. says seventy-three houses were utterly
destroyed from the time of Hugh the Sheriff. I., 75.

[384] Edred granted "to the religious woman, Elfthryth," supposed to
be the Abbess of Shaftesbury, "pars telluris Purbeckinga," which would
include Corfe. _Mon. Ang._, ii., 478.

[385] Both these kings spent large sums on Corfe Castle. See the
citations from the _Pipe Rolls_ in Hutchins' _Dorset_, vol. i., and in
Mr Bond's _History of Corfe Castle_.

[386] See Professor Baldwin Brown's paper in the _Journal of the
Institute of British Architects_, Third Series, ii., 488, and Mr
Micklethwaite's in _Arch. Journ._, liii., 338; also Professor Baldwin
Brown's remarks on Corfe Castle in _The Arts in Early England_, ii., 71.

[387] There are other instances in which the chapel is the oldest piece
of mason-work about the castle, as, for example, at Pontefract.

[388] Cited in Hutchins' _Dorset_, i., 488, from the _Close Rolls_.

[389] _Close Rolls_, i., 178b.

[390] Hutchins' _Dorset_, i., 488.

[391] Castrum Doveram, studio atque sumptu suo communitum. P. 108.
Eadmer makes Harold promise to William "Castellum Dofris cum puteo aquæ
ad opus meum te _facturum_." _Hist. Novorum_, i., d. The castle is not
mentioned in Domesday Book.

[392] _Norman Conquest_, iii., 217.

[393] In 1580 an earthquake threw down a portion of the cliff on which
the castle stands, and part of the walls. Statham's _History of Dover_,
p. 287.

[394] "Wendon him tha up to thære burge-weard, and ofslogen ægther
ge withinnan ge withutan, ma thanne 20 manna." Another MS. adds "tha
burh-menn ofslogen 19 men on othre healfe, and ma gewundode, and
Eustatius atbærst mid feawum mannum."

[395] See _ante_, pp. 17-19.

[396] His description is worth quoting:

    Est ibi mons altus, strictum mare, litus opacum,
    Hinc hostes citius Anglica regna petunt;
    Sed castrum Doveræ, pendens a vertice montis,
    Hostes rejiciens, littora tuta facit.
    Clavibus acceptis, rex intrans moenia castri
    Præcepit Angligenis evacuare domos;
    Hos introduxit per quos sibi regna subegit,
    Unumquemque suum misit ad hospitium.

"Carmen de Bello Hastingensi," in _Monumenta Britannica_, p. 603.

[397] William's description is also of great interest: "Deinde
dux contendit Doueram, ubi multus populus congregatus erat, pro
inexpugnabile, ut sibi videtur, munitione; quia id castellum situm est
in rupe mari contigua, quæ naturaliter acuta undique ad hoc ferramentis
elaborate incisa, in speciem muri directissima altitudine, quantum
sagittæ jactus permetiri potest, consurgit, quo in latere unda marina
alluitur." P. 140.

[398] The following entries in the _Pipe Rolls_ refer to this:--

 1194-5.  Three hundred planks of oak for the works of the
            castle                                               £2  0 0
 1196-7.  Repair of the wall of the castle                       76  3 0
 1208-9.  Timber for walling the castles of Dover and
            Rochester, also rods and [wooden] hurdles and
            other needful things                                 76 13 4
 1210-11. Payment for the carpenters working the timber          24  9 5
 1212-13. For the carriage of timber and other things            48 16 7
 1214-15. For the carriage of timber for the castle works         2  0 0
 1214-15. For timber and brushwood for the works, and for
            cutting down wood to make hurdles, and sending
            them                                          sum not given,

but £100 entered same year for the works of the castle. There is no
mention of stone for the castle during these two reigns, but after the
death of John we find that works are going on at Dover for which kilns
are required. (_Close Rolls_, i., 352, 1218.) This entry is followed by
a very large expenditure on Dover Castle (amounting to at least £6000),
sufficient to cover the cost of a stone wall and towers round the outer
circuit. The orders of planks for joists must be for the towers, and
the large quantities of lead, for roofing them. The order for timber
"ad palum et alia facienda" in 1225 _may_ refer to a stockade on the
advanced work called the Spur, which is said to be Hubert's work.
(_Close Rolls_, ii., 14.)

[399] Cited by Statham, _History of Dover_, pp. 265, 313.

[400] _Commune of London_, pp. 278-81.

[401] The ninth name, Maminot, is attached to three towers on the
curtain of the keep ward.

[402] "Recepto castro, quæ minus erant per dies octo addidit
firmamenta." P. 140.

[403] Lyon says: "The keep [hill] was formed of chalk dug out of the
interior hill." Cited by Statham, p. 245.

[404] "Per præceptum regis facta est apud Doveram turris fortissima."
II. 8, R. S., anno 1187. The _Historia Fundationis_ of St Martin's
Abbey says that Henry II. built the high tower in the castle, and
enclosed the donjon with new walls: "fit le haut tour en le chastel, et
enclost le dongon de nouelx murs." _M. A._, iv., 533.

[405] Puckle's _Church and Fortress of Dover Castle_, p. 57.

[406] _Pipe Rolls_, 1178-80. "In operatione muri circa castellum de
Doura, £165, 13s. 4d. The same, £94, 7s. 1d."

[407] Mr Statham thinks the port of Dover, though a Roman station, was
unwalled till the 13th century, and gives evidence. _History of Dover_,
p. 56.

[408] See Professor Baldwin Brown, "Statistics of Saxon Churches" in
the _Builder_, 20th October 1900; and in _The Arts in Early England_,
ii., 338.

[409] D. B., i., 1.

[410] "Istedem Willelmus tenet Dudelei, et ibi est castellum ejus. T.
R. E. valebat 4 libras, modo 3 libras." D. B., i., 177.

[411] _M. M. A._, i., 24.

[412] "Circa dies istos castellum de Huntinduna, de Waletuna, de
Legecestria, et Grobi, de Stutesbers [Tutbury], de Dudeleia, de Tresc,
et alia plura pariter corruerunt, in ultionem injuriarum quas domini
castellorum regi patri frequenter intulerunt." _Diceto_, i., 404, R. S.

[413] _Close Rolls_, i., 380.

[414] Parker's _History of Domestic Architecture_, Licenses to
Crenellate, 13th century, Part ii., p. 402. Godwin, "Notice of the
Castle at Dudley," _Arch. Journ._, xv., 47.

[415] D. B., i., 95b.

[416] Narrow terraces of this kind are found in several mottes, such
as Mere, in Wilts. They are probably natural, and may have been
utilised as part of the plan. The more regular terraces winding round
the motte are generally found where the motte has become part of a
pleasure-ground in later times.

[417] This is the only case in which I have had to trust to Mr Clark
for the description of a castle. _M. M. A._, ii., 24.

[418] Mentioned in _Close Rolls_, i., 518a.

[419] D. B., i., 95b.

[420] Symeon of Durham, 1072. "Eodem tempore, scilicet quo rex reversus
de Scotia fuerat, in Dunelmo castellum _condidit_, ubi se cum suis
episcopus tute ab incursantibus habere potuisset."

[421] This chapel is an instance of the honour so frequently done to
the chapel, which was in many cases built of stone when the rest of
the castle was only of timber, and was always the part most lavishly
decorated.

[422] The bailey was twice enlarged by Bishops Flambard and Pudsey.

[423] Surtees, Durham, iv., 33.

[424] Surtees Society, xx., 11-13.

[425] Evidently the southern wing wall up the motte; but we need not
suppose _murus_ to mean a stone wall.

[426] _Domus_, a word always used for a _habitation_ in mediæval
documents, and often applied to a tower, which it evidently means here.

[427] This is the only indication which Lawrence gives that the keep
was of wood.

[428]

    "Cingitur et pulchra paries sibi quilibet ala,
    Omnis et in muro desinit ala fero."

The translation is conjectural, but _gallery_ seems to make the best
sense, and the allusion probably is to the wooden galleries, or
_hourdes_, which defended the walls.

[429] Evidently the northern wing wall.

[430] This is the bailey; the two vast palaces must mean the hall
and the lodgings of the men-at-arms, who did not share the bishop's
dwelling in the keep. These were probably all of wood, as the buildings
of Durham Castle were burnt at the beginning of Pudsey's episcopate
(1153) and restored by him. Surtees Society, ix., 12.

[431] "Hujus in egressu pons sternitur." This seems a probable allusion
to a drawbridge, but if so, it is an early one.

[432] This describes the addition to the bailey made by Flambard. The
part of the peninsula to the S. of the church was afterwards walled in
by Pudsey, and called the South Bailey.

[433] _Liber Eliensis_, ii., 245 (Anglia Christiana). The part cited
was written early in the 12th century: see Preface.

[434] Stowe's _Annals_, 145, 1.

[435] D. B., ii., 192.

[436] "Alured de Merleberge tenet castellum de Ewias de Willelmo rege.
Ipse rex enim concessit ei terras quas Willelmus comes ei dederat, qui
hoc castellum refirmaverat, hoc est, 5 carucatas terræ ibidem.... Hoc
castellum valet 10_l_." D. B., i., 186a. As there is no statement of
the value in King Edward's day, we cannot tell whether it had risen or
fallen.

[437] _Feudal England_, p. 324. The present writer was led independently
to the same conclusion. Pentecost was the nickname of Osbern, son of
Richard Scrob, one of Edward's Norman favourites, to whom he had given
estates in Herefordshire. Osbern fled to Scotland in 1052, but he
seems to have returned, and was still holding lands in "the castelry
of Ewias" at the time of the Survey, though his nephew Alured held the
castle. See Freeman, _N. C._, ii., 345, and _Florence of Worcester_,
1052.

[438] "Locum vero intra moenia ad extruendum castellum delegit, ibique
Baldwinum de Molis, filium Gisleberti comitis, aliosque milites
præcipuos reliquit, qui necessarium opus conficerent, præsidioque
manerunt." Ordericus, ii., 181.

[439] Exeter is one of the few cities where a tradition has been
preserved of the site of the Saxon royal residence, which places it in
what is now Paul Street, far away from the present castle. Shorrt's
_Sylva Antiqua Iscana_, p. 7.

[440] "In hac civitate vastatæ sunt 48 domi postquam rex venit in
Angliam." D. B., i., 100.

[441] _Norman Conquest_, iv., 162.

[442] The outer ditch may have been of Roman origin, but in that case
it must have been carried all round the city, and we are unable to find
whether this was the case or not. The banks on the north and east sides
must also have been of Roman origin, and if we rightly understand the
statements of local antiquaries, the Roman city wall stood upon them,
and has actually been found _in situ_, cased with mediæval rubble.
_Report of Devon Association_, 1895.

[443] This resemblance to a pit may be seen in every motte which still
retains its ancient earthen breast-work, as at Castle Levington, Burton
in Lonsdale, and Castlehaugh, Gisburne. Perhaps this is the reason
that we so frequently read in the _Pipe Rolls_ of "the houses _in_ the
motte" (domos in Mota) instead of _on_ the motte. Devizes Castle is
another and still more striking instance.

[444] Professor Baldwin Brown, _The Arts in Early England_, ii., 82.

[445] "In custamento gaiole in ballia castelli, £16, 15s. 8d."

[446] Cited by Dr Oliver, "The Castle of Exeter," in _Arch. Journ._,
vii., 128.

[447] The whole of this passage is worth quoting: "Castellum in ea
situm, editissimo aggere sublatum, muro inexpugnabile obseptum,
turribus Cæsarianis inseissili calce confectis firmatum. Agmine
peditum instructissime armato exterius promurale, quod ad castellum
muniendum aggere cumulatissimo in altum sustollebatur, expulsis
constanter hostibus suscepit, pontemque interiorem, quo ad urbem de
castello incessus protendebatur, viriliter infregit, lignorumque
ingentia artificia, quibus de muro pugnare intentibus resisteretur,
mire et artificiose exaltavit. Die etiam et noctu graviter et intente
obsidionem clausis inferre; nunc cum armatis aggerem incessu quadrupede
conscendentibus rixam pugnacem secum committere; nunc cum innumeris
fundatoribus, qui e diverso conducti fuerunt, intolerabile eos lapidum
grandine infestare; aliquando autem ascitis eis, qui massæ subterranæ
cautius norunt venus incidere, ad murum diruendum viscera terræ scutari
præcipere: nonnunquam etiam machinas diversi generis, alias in altum
sublatis, alias humo tenus depressas, istas ad inspiciendam quidnam
rerum in castello gereretur, illas ad murum quassandum vel obruendum
aptare." _Gesta Stephani_, R. S., 23.

[448] _Pipe Rolls_, 1169-1186.

[449] The difficulty about this, however, is that passages branch off
from the central cave in every direction.

[450] Oliver's _History of Exeter_, p. 186.

[451] [Willelmus Malet] fecit suum castellum ad Eiam. D. B., ii., 379.
For Malet, see Freeman, _N. C._, 466, note 4.

[452] "In operatione castelli de Eya et reparatione veterarum
bretascharum et 2 novarum bretascharum et fossatorum et pro carriagio
et petra et aliis minutis operationibus 20_l._ 18_s._ 4_d._" _Pipe
Rolls_, xix., 19 Henry II. The small quantity of stone referred to here
can only be for some auxiliary work. The _bretasches_ in this case will
be mural towers of wood. "In emendatione palicii et 1 exclusæ vivarii
et domorum castelli 20_s._" 28 Henry II.

[453] D. B., ii., 319, 320.

[454] D. B., i., 162. "Sedecim domus erant ubi sedet castellum, quæ
modo desunt, et in burgo civitatis sunt wastatæ 14 domus."

[455] Rudge, _History of Gloucester_, p. 7. Haverfield, _Romanisation
of Britain_, p. 204.

[456] It is, however, possible that by the _burgus_ may be meant a
later quarter which had been added to the city.

[457] Fosbroke's _History of Gloucester_, pp. 125, 126. Stukeley,
writing in 1721, says: "There is a large old gatehouse standing, and
near it the castle, with a very high artificial mount or keep nigh the
river." _Itin. Cur._, i., 69.

[458] "Of al partes of yt the hy tower _in media area_ is most
strongest and auncient." Leland, _Itin._, iii., 64.

[459] "In excambium pro placea ubi nunc turris stat Gloucestriæ, ubi
quondam fuit ortus monachorum." _Mon. Ang._, i., 544. The document is
not earlier than Henry II.'s reign.

[460] Round, _Studies in Domesday_, p. 123.

[461] "In operatione frame turris de Glouec, 20_l._" _Pipe Rolls_,
i., 27. In the single _Pipe Roll_ of Henry I. there is an entry "In
operationibus turris de Glouec," 7_l._ 6_s._ 2_d._, which _may_ be one
of a series of sums spent on the new stone keep.

[462] _Pipe Rolls_, 1177, 1180, 1181, 1184.

[463] _Close Rolls_, ii., 88b.

[464] "In reparatione murorum et bretaschiarum," 20_l_. 7_s._ 11_d._
_Pipe Rolls_, 1193.

[465] "Jussit ut foderetur castellum ad Hestengaceastra."

[466] D. B., i., 18a, 2. "Rex Willelmus dedit comiti [of Eu]
castellariam de Hastinges."

[467] "Dux ibidem [at Pevensey] non diu moratus, haud longe situm, qui
Hastinges vocatur, cum suis adiit portum, ibique opportunum nactus
locum, ligneum agiliter castellum statuens, provide munivit." _Chron.
Monast. de Bello_, p. 3, ed. 1846. There is also the evidence of
Ordericus, who says that Humphrey de Tilleul received the custody of
Hastings Castle "from the first day it was built." iv., 4.

[468]

    Par conseil firent esgarder
    Boen lieu a fort chastel fermer.
    Donc ont des nes mairrien iete,
    A la terre l'ont traine,
    Que le quens d'Ou i out porte
    Trestot percie e tot dole.
    Les cheuilles totes dolees
    Orent en granz bariz portees.
    Ainz que il fust avespre
    En ont un chastelet ferme;
    Environ firent une fosse,
    Si i ont fait grant fermete.--Andresen's edition, p. 289.

[469] The north curtain is of ruder work than the other masonry.

[470] In attractu petre et calcis ad faciendam turrim de Hasting
6_l._ Idem 13_l._ 12_s._ Vol. xviii., p. 130. The work must have been
extensive, as it is spoken of as "operatio castelli novi Hasting."
1181-1182. Though the sum given is not sufficient for a great stone
keep, it may have been supplemented from other sources.

[471] See Mr Sands' paper on Hasting's Castle, in _Trans. of the
South-Eastern Union of Scientific Societies_, 1908.

[472] This bailey has been supposed to be a British or Roman earthwork,
but no evidence has been brought forward to prove it, except the fact
that discoveries made in one of the banks point to a flint workshop on
the site.

[473] Totum manerium valebat T. R. E. 20 libras, et postea wastum fuit.
Modo 18 libras 10 solidos. D. B., i., 18a, 2.

Since the above was written, Mr Chas. Dawson's large and important work
on Hastings Castle has appeared, and to this the reader is referred
for many important particulars, especially the passages from the _Pipe
Rolls_, i., 56, and the repeated destructions by the sea, ii., 498-9.
The reproduction of Herbert's plan of 1824 (ii., 512) seems to show
more than one bailey outside the inner ward. The evidence for a great
outer ditch, enclosing all these works, and supposed to be prehistoric,
is given on p. 515, vol. ii.

[474] See _Anglo-Saxon Chronicle_, 1048 (Peterborough) and 1052
(Worcester), and compare with _Florence of Worcester_.

[475] _N. C._, ii., 394.

[476] _Pipe Rolls_, 11 Henry II., p. 100, and 15 Henry II., p. 140.
Stephen granted to Miles of Gloucester "motam Hereford cum toto
castello." Charter cited by Mr Round, _Geoffrey de Mandeville_,
Appendix O, p. 329.

[477] Cited by Grose, _Antiquities_, ii., 18. Stukeley saw the motte,
and mentions the well in it lined with stone. _Itin. Curiosum_, i., 71.
See also Duncombe's _History of Hereford_, i., 229.

[478] In custamento prosternandi partem muri castri nostri de Hereford,
et preparatione rogi ad reficiendum predictum murum, 26s. 6d. _Pipe
Rolls_, 1181-1182.

[479] In operatione 5 bretaschiarum in castro de Hereford, £15, 3s. 9d.
_Pipe Rolls_, 1173-1174.

[480] _Close Rolls_, i., 134a.

[481] Hubertus Cantuariensis Archiepiscopus et totius Angliæ summus
Justiciarius, fuit in Gwalia apud Hereford, et recepit in manu sua
castellum de Hereford, et castellum de Briges, et castellum de
Ludelaue, expulsis inde custodibus qui ea diu custodierant, et tradidit
ea aliis custodibus, custodienda ad opus regis. _Roger of Howden_, iv.,
35, R. S.

[482] D. B., i., 179.

[483] "In loco castri fuerunt 20 mansiones, quæ modo absunt." D. B.,
i., 203.

[484] _Ordericus_, ii., 185.

[485] _Benedict of Peterborough_, i., 70. The Justiciar, Richard de
Lucy, threw up a siege castle against it.

[486] "Pro uncis ad prosternandum palicium de Hunted, 7_s._ 8_d._ In
operatione novi castelli de Hunted, et pro locandis carpentariis et pro
croccis et securibus et aliis minutis rebus, 21_l._" _Pipe Rolls_, 20
Henry II., pp. 50, 63. It is clear that the _operatio_ was in this case
one of pulling down. Giraldus (_Vita Galfredi_, iv., 368, R. S.) and
_Diceto_ (i., 404, R. S.), both say the castle was destroyed.

[487] _Mon. Ang._, vi., 80.

[488] Leland tells us that Launceston was anciently called Dunheved.
_Itin._, vii., 122.

[489] "Ibi est castrum comitis." D. B., i., 121b. "Hæc duo maneria
[Hawstone et Botintone] dedit episcopo comes Moriton pro excambio
castelli de Cornualia." D. B., i., 101b, 2.

[490] There are no entries for Launceston except repairs in the reigns
of Henry II. and his sons.

[491] Murray's _Guide to Cornwall_, p. 203.

[492] "Olim 20_l._; modo valet 4_l._" D. B., i., 121b.

[493] D. B., ii., 157, 163, 172. The first entry relating to this
transaction says: "Hoc totum est pro escangio de 2 maneriis Delaquis."
The second says: "Pertinent ad castellum Delaquis." It is clear that
Lewes is meant, as one paragraph is headed "De escangio Lewes." I have
been unable to find any explanation of this exchange in any of the
Norfolk topographers, or in any of the writers on Domesday Book.

[494] Lincoln is the only other instance known to the writer. Deganwy
has two natural mottes. It is possible that two mottes indicate a
double ownership of a castle, a thing of which there are instances, as
at Rhuddlan.

[495] Exeter and Tickhill are instances of early Norman gateways, and
at Ongar and Pleshy there are fragments of early gateways, though there
are no walls on the banks. We have already seen that Arundel had a
gateway which cannot be later than Henry I.'s time.

[496] D. B., i., 26a, 1.

[497] "De predictis wastis mansionibus propter castellum destructi
fuerunt 166." D. B., i., 336b, 2.

[498] "In reversione sua Lincoliæ, Huntendonæ, et Grontebrugæ castra
locavit." Ordericus, 185 (Prévost).

[499] At present the bank is wanting on a portion of the south side,
between the two mottes.

[500] Mr Clark gravely argues that the houses were inside what he
believes to have been the Saxon castle. There is not a vestige of
historical evidence for the existence of any castle in Lincoln in the
Saxon period.

[501] Stephen gave Ralph the castle and city of Lincoln, and gave him
leave to fortify one of the towers in Lincoln Castle, and have command
of it until the king should deliver to him the castle of Tickhill; then
the king was to have the city and castle of Lincoln again, excepting
the earl's own tower, which his mother had fortified. His mother was
Lucy, daughter of Ivo Taillebois; and as the principal tower was known
as the Luce Tower, the masonry may have been her work. In that case the
Norman work on the smaller motte may be due to Ralph Gernon, and may
possibly be the _nova turris_ which was repaired in John's reign. _Pipe
Roll_, 2 John. Stephen's charter is in Farrer's _Lancashire Pipe Rolls_.

[502] "In custamento firmandi ballium castelli Lincoll." _Pipe Roll_,
5 Richard I. In an excavation made for repairs in modern times it was
found that this wall rested on a timber frame-work, a device to avoid
settling, the wall being of great height and thickness. Wilson, Lincoln
Castle, _Proc. Arch. Inst._, 1848.

[503] D. B., i. 336b, 2: "Tochi filius Outi habuit in civitate 30
mansiones præter suam hallam, et duas ecclesias et dimidiam, et suam
hallam habuit quietam ab omni consuetudine.... Hanc aulam tenuit
Goisfredus Alselin et suus nepos Radulfus. Remigius episcopus tenet
supradictas 30 mansiones ita quod Goisfredus nihil inde habet."

[504] "In castello Monemouth habet Rex in dominio 4 carucas. Willelmus
filius Baderon custodit eas. Quod rex habet in hoc castello valet c
solidos." D. B., 180b.

[505] _Liber Landavensis_, Evans' edition, pp. 277-278. See also
Round's _Calendar of Documents Preserved in France_, p. 406.

[506] _Theatre of Britain_, p. 107.

[507] Speed's map shows the curtain wall surrounding the top of the
hill and also a large round tower towards the N.E. part, but not
standing on any "other mount." The square keep is not indicated
separately. It must be remembered that Speed's details are not always
accurate or complete.

[508] "Ipse comes tenet in dominio Bishopstowe, et ibi est castellum
ejus quod vocatur Montagud. Hoc manerium geldabat T. R. E. pro 9 hidas,
et erat de abbatia de Adelingi, et pro eo dedit comes eidem ecclesiæ
manerium quod Candel vocatur." D. B., i., 93a, 1.

[509] _Itin._, ii., 92.

[510] From a description communicated by Mr Basil Stallybrass. The
motte is shown in a drawing in Stukeley's _Itinerarium Curiosum_. The
"immense Romano-British camp" of which Mr Clark speaks (_M. M. A._, i.,
73) is nearly a mile west.

[511] Mountjoy, Monthalt (Mold), Beaumont, Beaudesert, Egremont, are
instances in point.

[512] _Gaimar_, 214, Wright's edition. Gaimar wrote in the first half
of the 12th century; Wright states that his work is mainly copied
from the _Anglo-Saxon Chronicle_, but its chief value lies in the old
historical traditions of the north and east of England which he has
preserved.

[513] Hodgson's _History of Northumberland_, Part II., ii., 384, 389.

[514] This account is taken from a description kindly furnished by Mr
D. H. Montgomerie.

[515] Bates' _Border Holds_, p. 11.

[516] _Simeon of Durham_, 1080. "Castellum Novum super flumen Tyne
condidit."

[517] See the map in an important paper on Newcastle by Longstaffe,
_Arch. Æliana_, iv., 45.

[518] _Guide to the Castle of Newcastle_, published by Society of
Antiquaries of Newcastle, 1901.

[519] Longstaffe, as above.

[520] "Condidit castellum in excelso preruptæ rupis super Twedam
flumen, ut inde latronum incursus inhiberet, et Scottorum irruptiones.
Ibi enim utpote in confinia regni Anglorum et Scottorum creber
prædantibus ante patebat excursus, _nullo enim quo hujusmodi impetus
repelleretur præsidio locato_." _Symeon of Durham_, R. S., i., 140.

[521] "Castellum di Northam, quod munitionibus infirmum reperit, turre
validissima forte reddidit." _Geoffrey of Coldingham_, 12 (Surtees
Society). Symeon says it was built "precepto regis." The keep was
extensively altered in the Decorated period.

[522] _M. M. A._, ii., 331.

[523] _Richard of Hexham_, 319 (Twysden).

[524] "In illa terra de quâ Herold habebat socam sunt 15 burgenses
et 17 mansuræ vastæ, quæ sunt in occupatione castelli; et in burgo
190 mansuræ vacuæ in hoc quod erat in soca regis et comitis, et 81 in
occupatione castelli." D. B., ii., 116. This shows that the castle
and its ditches occupied ground partly within and partly without the
ancient _burh_.

[525] Harrod's _Gleanings among Castles_, p. 142.

[526] The authorities from which this map is compiled are not given.

[527] The "new borough" at Norwich was the quarter inhabited by
the Normans. D. B., ii., 118. "Franci de Norwich: in novo burgo 36
burgenses et 6 Anglici." Mr Hudson says that Mancroft Leet corresponds
to the new burgh added to Norwich at the Conquest. See his map in
_Arch. Journ._, xlvi.

[528] Norwich was not a Roman town; see Haverfield, _Vict. Hist. of
Norfolk_, i., 320. But the Roman road from Caistor passed exactly
underneath the castle motte. _Brit. Arch. Assoc. Journ._, xlvi., Rev.
H. Dukinfield Astley.

[529] Harrod's _Gleanings among Castles_, p. 137.

[530] _Mon. Ang._, iv., 13. In 37 Henry III. the monks of Norwich
Priory received "licentiam includendi eandem villam cum fossis," and by
doing this they enclosed the lands of other fees.

[531] _Arch. Journ._, xlvi., 445.

[532] Kirkpatrick's _Notes of Norwich Castle_, written about 1725. He
states that the angles of the motte had been spoilt, and much of it
fallen away.

[533] _Archæologia_, vol. xii.

[534] Mr Hartshorne thought it was built between 1120 and 1125. _Arch.
Journ._, xlvi., 260. It is certainly not as late as Henry II.'s reign,
or the accounts for it would appear in the _Pipe Rolls_.

[535] _Pipe Rolls_, 19 Henry II., p. 117. In reparatione pontis lapidei
et palicii et 3 bretascharum in eodem castello, 20_l._ 4_s._ 8_d._

[536] _Close Rolls_, ii., 22. Order that the palicium of Norwich
Castle, which has fallen down and is threatened with ruin, be repaired.

[537] Kirkpatrick, _Notes on Norwich Castle_.

[538] Except Kirkpatrick, who shows a judicious scepticism on the
subject. _Ibid._, p. 248.

[539] _Mon. Ang._, i., 482.

[540] D. B., ii., 117.

[541] Ordericus, ii., 184.

[542] Published in a paper on Nottingham Castle by Mr Emanuel Green, in
_Arch. Journ._ for December 1901.

[543] See Mr Green's paper, as above, p. 388.

[544] "Apud Rokingham liberavimus Philippo Marco ad faciendam turrim
quam dominus Rex precepit fieri in Mota de Notingham 100 marcas quas
burgenses de Notingham et Willelmus Fil. Baldwini dederunt domino Regi
pro benevolencia sua habenda." In Cole's _Documents Illustrative of
English History_, 235. There is some reason to think that John instead
of building the cylindrical keeps which were then coming into fashion,
reverted to the square form generally followed by his father.

[545] _Pipe Rolls_, 1170-1186. The _Pipe Roll_ of 6 Richard I. mentions
the making of "1 posterne in mota," which may be the secret passage in
the rock.

[546] This is rendered probable by a writ of Henry III.'s reign,
ordering that half a mark is to be paid annually to Isolde de Gray
for the land which she had lost in King John's time "_per incrementum
forinseci ballii Castri de Notinge_." _Close Rolls_, i., 508.

[547] _Close Rolls_, i., 548b. "Videat quid et quantum mæremii opus
fuerit ad barbecanas et palitia ipsius castri reparanda" (1223).
_Close Rolls_, i., 531b. Timber ordered for the repair of the bridges,
bretasches, and _palicium gardini_ (1223). _Cal. of Close Rolls_, 1286,
p. 390: Constable is to have timber to repair the weir of the mill, and
the _palings of the court_ of the castle. Nottingham was one of eight
castles in which John had baths put up. _Rot. Misæ._, 7 John.

[548] The murage of the town of Nottingham was assigned "to the repair
of the outer bailey of the castle there" in 1288. _Patent Rolls_,
Edward I., i., 308.

[549] Chapter xlii.

[550] D. B., i., 280.

[551] "Ipse Baldwinus vicecomes tenet de Rege Ochementone, et ibi sedet
castellum." D. B., i., 105b, 2.

[552] The late Mr Worth thought the lower part of the keep was early
Norman. He was perhaps misled by the round arched loops in the
basement. But round arches are by no means conclusive evidence in
themselves of Norman date, and the size of these windows, as well as
the absence of buttresses, and the presence of pointed arches, are
quite incompatible with the early Norman period. The whole architecture
of the castle agrees with a 14th century date, to which the chapel
undoubtedly belongs.

[553] Eyton, _Antiquities of Shropshire_, vol. vii.

[554] "Ibi fecit Rainaldus Castellum Luure." D. B., i., 253b. Rainald
was an under-tenant of Roger, Earl of Shrewsbury.

[555] This sketch is reproduced in Mr Parry-Jones' _Story of Oswestry
Castle_. Leland says, "Extat turris in castro nomine Madoci." _Itin._,
v., 38.

[556] "In operatione palicii de Blancmuster 2_l._ 6_s._ 8_d._" XII.,
124. Oswestry was known as Blancmoustier or Album Monasterium in Norman
times.

[557] _Abingdon Chronicle_ and _Osney Chronicle_, which, though both of
the 13th century, were no doubt compiled from earlier sources.

[558] _Osney Chronicle_, 1071.

[559] See Ingram's _Memorials of Oxford_ for an account of the very
interesting crypt of this church, p. 8. The battlement storey of the
tower is comparatively late.

[560] Mackenzie, _Castles of England_, i., 160.

[561] D. B., p. 154.

[562] Rymer's _Foedera_, vol. i.

[563] "Terram castelli Pechefers tenuerunt Gerneburn et Hunding." D.
B., i., 276a, 2.

[564] There are similar nook-shafts to Henry II.'s keep at Scarborough,
and to Castle Rising. Mr Hartshorne (_Arch. Journ._, v., 207) thought
that there had been an earlier stone keep at Peak Castle, because
some moulded stones are used in the walls, and because there is some
herring-bone work in the basement. But this herring-bone work only
occurs in a revetment wall to the rock in the cellar; and the moulded
stones may be quite modern insertions for repairs, and may have come
from the oratory in the N.E. angle, or from some of the ruined windows
and doorways. The sums entered to this castle between the years 1172
and 1176 are less than half the cost of Scarborough keep, and do not
appear adequate, though the keep was a small one. But there is some
reason to think that the cost of castles was occasionally defrayed in
part from sources not entered in the _Pipe Rolls_.

[565] Rex E. tenuit Peneverdant. Ibi 2 carucatæ terræ et reddebant 10
denarios. Modo est ibi castellum.... Valent 3 libras. D. B., i., 270.

[566] We need not resort to any fanciful British origins of the name
Peneverdant, as it is clearly the effort of a Norman scribe to write
down the unpronounceable English name Penwortham.

[567] See _ante_, under Clitheroe.

[568] Mr Halton's book (_Documents relating to the Priory of
Penwortham_) throws no light on this point.

[569] _Transactions of the Historic Society of Lancashire and
Cheshire_, vol. ix., 1856-1857, paper on "The Castle Hill of
Penwortham," by the Rev. W. Thornber; Hardwick's _History of Preston_,
pp. 103-11.

[570] In a paper published in the _Trans. Soc. Ant. Scot_, for 1900, on
"Anglo-Saxon Burhs and Early Norman Castles," the present writer was
misled into the statement that this hut was the remains of the cellar
of the Norman _bretasche_. A subsequent study of Mr Hardwick's more
lucid account of the excavations showed that this was an error. There
were two pavements of boulders, one on the natural surface of the hill,
on which the hut had been built, the other 5 feet above it, and 12 feet
below the present surface. The hut appeared to have been circular, with
wattled walls and a thatched roof. Several objects were found in its
remains, and were pronounced to be Roman or Romano-British. The upper
pavement would probably be the flooring of a Norman keep.

[571] Mr Roach Smith pronounced this spur to be Norman. As its evidence
is so important, it is to be regretted that its position was not more
accurately observed. It was found in the lowest stratum of the remains,
but Mr Hardwick says: "As it was not observed until thrown to the
surface, a possibility remained that it might have fallen from the
level of the upper boulder pavement, 5 feet higher." We may regard this
possibility as a certainty, if the lower hut was really British.

[572] Mr Willoughby Gardner says the castle commands a ford, to which
the ancient sunk road leads. _Victoria Hist. of Lancashire_, vol. ii.

[573] Hugh Candidus, _Coenob. Burg. Historia_, in Sparke's _Scriptores_,
p. 63. This passage was kindly pointed out to me by Mr Round. Hugh
lived in Henry III.'s reign, but he must have had the more ancient
records of the monastery at his disposal.

[574] Domesday Book mentions that the value of the burgus had greatly
risen. It was one of the _burhs_ mentioned in the _Burghal Hidage_.

[575] _Pipe Roll_, 1187-1188. William of Jumièges says, "Statim
firmissimo vallo castrum condidit, probisque militibus commisit." VII.,
34. Wace professes to give the account of an eye-witness, who saw
the timber for the castle landed from the ships, and the ditch dug.
But Wace was not a contemporary, and as he has made the mistake of
making William land at Pevensey instead of Hastings, his evidence is
questionable. _Roman de Rou_, p. 293 (Andresen's edition).

[576] The ruins of this keep, until 1908, were buried under so large a
mound of earth and rubbish that Mr G. T. Clark mistook it for a motte,
and the present writer was equally misled. It ought to be stated,
before the date of this keep is finally settled, that the _Gesta
Stephani_ speaks of this castle as "editissimo aggere sublatum." P. 106.

[577] _Ibid._

[578] _Close Rolls_, i., 631a.

[579] D. B., i., 20b.

[580] D. B., i., 373b.

[581] Cited in Holmes' _History of Pontefract_, p. 62.

[582] Another charter, which is a confirmation by the second Ilbert
de Lacy of the ecclesiastical gifts of Ilbert I. and Robert his son,
states that the Chapel of St Clement in the castle of Pontefract was
founded by Ilbert I. in the reign of William II. _Mon. Ang._, v., 128.

[583] It is not necessary to discuss the meaning of the name
Pontefract, since for whatever reason it was given, it was clearly
bestowed by the Norman settlers.

[584] "Castrum de Pontefracto est quasi clavis in comitatu Ebor."
Letter of Ralph Neville to Henry III., _Foedera_, i., 429, cited by
Holmes, _Pontefract_, 194.

[585] The Conqueror had given him more than 200 manors in Yorkshire.
_Yorks. Arch. Journ._, xiv., 17.

[586] Four roundels are shown in the plate given in Fox's _History
of Pontefract_, "from a drawing in the possession of the Society of
Antiquaries." But the drawing is so incorrect in some points that it
can hardly be relied upon for others. There were only three roundels in
Leland's time.

[587] Drake's account of the siege says that there was a hollow place
between Piper's Tower and the Round Tower all the way down to the well;
the gentlemen and soldiers all fell to carrying earth and rubbish, and
so filled up the place in a little space. Quoted in Holmes' _Manual of
Pontefract Castle_.

[588] In the _English Historical Review_ for July 1904, where this
paper first appeared, the writer spoke of _two_ mottes at Pontefract,
having been led to this view by the great height of the east end of the
bailey, where the ruins of John of Gaunt's work are found. This view is
now withdrawn, in deference to the conclusions of Mr D. H. Montgomerie,
F.S.A., who has carefully examined the spot.

[589] _Mon. Ang._, iv., 178.

[590] From a description by Mr D. H. Montgomerie.

[591] D. B., i., 224.

[592] See Chapter IV.

[593] Domesday Book says: "Ipse comes (Roger) tenet Ardinton. Sancta
Milburga tenuit T. R. E. Ibi molinum et nova domus et burgus Quatford
dictus, nil reddentes." I., 254.

[594] G. T. Clark, in _Arch. Cambrensis_, 1874, p. 264.

[595] _Ord. Vit._, iv., 32.

[596] "In hoc manerio fecit Suenus suum castellum." D. B., ii., 33b.

[597] Freeman, _N. C._, ii., 329, and iv., Appendix H.

[598] Mr Round has suggested that this castle was at Canfield in Essex,
where there is a motte and bailey.

[599] "Isdem Osbernus habet 23 homines in castello Avreton et reddit 10
solidos. Valet ei castellum hoc 20 solidos." D. B., i., 186b.

[600] Mr Clark's plan is strangely incorrect, as he altogether omits
the bailey. Compare the plan in Mr Round's Castles of the Conquest,
_Archæologia_, vol. lviii., and Mr Montgomerie's plan here, Fig. 27.

[601] "Comes Alanus habet in sua castellata 199 maneria.... Præter
castellariam habet 43 maneria." D. B., i., 381a, 2.

[602] This is stated in a charter of Henry II., which carefully
recapitulates the gifts of the different benefactors to St Mary's.
_Mon. Ang._, iii., 548. It is curious that the charter of William II.,
the first part of which is an inspeximus of a charter of William I.,
does not mention this chapel in the castle.

[603] Mr Skaife, the editor of the _Yorkshire Domesday_, thinks that
it was at Hinderlag, but gives no reasons. Hinderlag, at the time of
the Survey, was in the hands of an under-tenant. _Yorks. Arch. Journ._,
lii., 527, 530.

[604] "Hic Alanus primo incepit facere castrum et munitionem juxta
manerium suum capitale de Gilling, pro tuitione suorum contra
infestationes Anglorum tunc ubique exhæredatorum, similiter et Danorum,
et nominavit dictum castrum Richmond suo ydiomate Gallico, quod sonat
Latine divitem montem, in editiori et fortiori loco sui territorii
situatum." _Mon. Ang._, v., 574.

[605] There are no remains of fortification at Gilling, but about a
mile and a half away there used to be an oval earthwork, now levelled,
called Castle Hill, of which a plan is given in M'Laughlan's paper,
_Arch. Journ._, vol. vi. It had no motte. Mr Clark says, "The mound at
Gilling has not long been levelled." _M. M. A._, i., 23. It probably
never existed except in his imagination.

[606] See Clarkson's _History of Richmond_.

[607] _Journal of Brit. Arch. Ass._, lxiii., 179.

[608] These are the dates given in Morice's _Bretagne_.

[609] Henry spent 51_l._ 11_s._ 3_d._ in 1171 on "operationes domorum
et turris," and 30_l._ 6_s._ in 1174 on "operationes castelli et
domorum."

[610] "Episcopus de Rouecestre, pro excambio terræ in qua castellum
sedet, tantum de hac terra tenet quod 17 sol. et 4 den. valet." D. B.,
i., 2b.

[611] See Mr George Payne's paper on _Roman Rochester_, in _Arch.
Cantiana_, vol. xxi. Mr Hope tells me that parts of all the four sides
are left.

[612] Thus Egbert of Kent, in 765, gives "terram intra castelli
moenia supra-nominati, id est Hrofescestri, unum viculum cum duobus
jugeribus," _Kemble_, i., 138; and Offa speaks of the "episcopum
castelli quod nominatur Hrofescester," Earle, _Land Charters_, p. 60.

[613] See an extremely valuable paper on _Mediæval Rochester_ by the
Rev. Greville M. Livett, _Arch. Cantiana_, vol. xxi.

[614] See the charter of Coenulf, King of Mercia, giving to Bishop
Beornmod three ploughlands on the southern shore of the city of
Rochester, from the highway on the east to the Medway on the west.
_Textus Roffensis_, p. 96.

[615] The name Boley may possibly represent the Norman-French
_Beaulieu_, a favourite Norman name for a castle or residence.
Professor Hales suggested that Boley Hill was derived from Bailey
Hill (cited in Mr Gomme's paper on Boley Hill, _Arch. Cantiana_, vol.
xvii.). The oldest form of the name is Bullie Hill, as in Edward IV.'s
charter, cited below, p. 200.

[616] Roman urns and lachrymatories were found in the Boley Hill when
it was partially levelled in the 18th century to fill up the castle
ditch. _History of Rochester_, p. 281. At the part now called Watt's
Avenue, Mr George Payne found "the fag-end of an Anglo-Saxon cemetery."
_Arch. Cantiana_, vol. xxi.

[617] "In pulchriore parte civitatis Hrouecestre." _Textus Roffensis_,
p. 145. Mr Freeman and others have noticed that the special mention of
a _stone_ castle makes it probable that the first castle was of wood.
Mr Round remarks that the building of Rochester Castle is fixed, by the
conjunction of William II. and Lanfranc in its history, to some date
between September 1087 and March 1089. _Geoffrey de Mandeville_, p.
339. Probably, therefore, it was this new castle which Bishop Odo held
against Rufus in 1088. Ordericus says that "cum quingentis militibus
intra Rofensem urbem se conclusit." P. 272.

[618] It is now attributed to Archbishop William of Corbeuil, to whom
Henry I. gave the custody of the castle in the twenty-seventh year of
his reign, with permission to make within it a defence or keep, such as
he might please. _Continuator of Florence_, 1126. Gervase of Canterbury
also says "idem episcopus turrim egregiam ædificavit." Both passages
are cited by Hartshorne, _Arch. Journ._, xx., 211. Gundulf's castle
cost 60_l._ and can scarcely have been more than an enclosing wall with
perhaps one mural tower. See Mr Round, _Geoffrey de Mandeville_, 340,
and Mr Livett's paper, cited above.

[619] Two common friends of Rufus and Gundulf advised the king that
in return for the grant of the manor of Hedenham and the remission
of certain moneys, "episcopus Gundulfus, quia in opere cæmentario
plurimum sciens et efficax erat, castrum sibi Hrofense _lapideum_ de
suo construeret." _Textus Roffensis_, p. 146. There was therefore an
exchange of land in this affair also.

[620] _Arch. Cantiana_, vol. xxi.

[621] _Arch. Cantiana_, vol. xxi., p. 49.

[622] There are several entries in the _Close Rolls_ relating to this
wall of Henry III. in the year 1225.

[623] Mr Beale Poste says that this ancient wall was met with some
years since in digging the foundations of the Rev. Mr Conway's house,
standing parallel to the present brick walls and about 2 feet within
them. "Ancient Rochester as a Roman Station," _Arch. Cantiana_, ii.,
71. The Continuator of Gervase of Canterbury tells us (ii., 235) that
at the siege of Rochester in 1265, Simon de Montfort captured the outer
castle up to the keep (forinsecum castellum usque ad turrim), and Mr
Livett thinks this outer castle must have been the Boley Hill.

[624] _Close Rolls_, ii., 98b.

[625] Hasted's _Kent_, iv., 163.

[626] "Ymb sætan tha ceastre and worhton other fæsten ymb hie selfe."
See _ante_, p. 49, _note 120_.

[627] Mr Hope suggests the east side, as the north was a marsh.

[628] _History of Rochester_ (published by Fisher, 1772), p. 285.

[629] D. B., i., 56.

[630] "Wasta erat quando Rex W. iussit ibi castellum fieri. Modo valet
36 solidos." D. B., i., 220.

[631] "I markid that there is stronge Tower in the Area of the
Castelle, and from it over the Dungeon Dike is a drawbridge to the
Dungeon Toure." _Itin._, i., 14.

[632] "In operatione nove turris et nove camere in cast. 126_l._ 18_s._
6_d._"

[633] D. B., i., 120.

[634] See the plan reproduced in Wise's _Rockingham Castle and the
Watsons_, p. 66.

[635] Vol. i., p. 224: cited by Mr Irving in his valuable paper on Old
Sarum in _Arch. Journ._, xv., 1859. Sir Richard made a vague reference
to an MS. in the Cottonian and Bodleian libraries, for which Mr Irving
says he has searched in vain.

[636] General Pitt-Rivers in his Address to the Salisbury meeting
of the Archæological Institute in 1887, says that traces of these
roads may still be seen. He adds that Old Sarum does not resemble the
generality of ancient British fortifications, in that the rampart is
of the same height all round, instead of being lower where the ground
is steeper; this led him to think that the original fortress had been
modernised in later times. Sir Richard Colt Hoare noticed that the
ramparts of Sarum were twice as high as those of the fine prehistoric
camps with which he was acquainted. _Ancient Wiltshire_, p. 226.

[637] Benson and Hatcher's _Old and New Sarum_, p. 604.

[638] _Cf._ Benson and Hatcher, 63, with _Beauties of England and
Wales_, xv., 78.

[639] D. B., i., 66. "Idem episcopus tenet Sarisberie." Part of the
land which had been held under the bishop was now held by Edward the
Sheriff, the ancestor of the earls of Salisbury. This in itself is a
proof that the castle was new. See Freeman, _N. C._, iv., 797.

[640] This policy had been dictated by an oecumenical council.

[641] He gives to the canons of the church two hides in the manor,
"et ante portam castelli Seriberiensis terram ex utraque parte viæ in
ortorum domorumque canonicorum necessitate." _M. A._, vi., 1294.

[642] _Gentleman's Magazine_, 1795.

[643] The area of the outer camp is 29-1/2 acres.

[644] It is unlikely that this is the _turris_ mentioned in the
solitary _Pipe Roll_ of Henry I. "In unum ostium faciendum ad cellarium
turris Sarum, 20s." This entry is of great interest, as entrances
from the outside to the basement of keeps were exceptional in the
12th century; but the basement entrance of Colchester keep has every
appearance of having been added by Henry I.

[645] William of Malmesbury, _Hist. Nov._, ii., 91.

[646] In 1152; the writ is given by Benson and Hatcher, p. 32.

[647] "In operatione unius Bretesche in eodem Castro 50s." _Pipe
Rolls_, 1193-4.

[648] "Virgam et mairemium ad hordiandum castrum." _Close Rolls_, i.,
198b (1215).

[649] Benson and Hatcher, p. 704.

[650] "Dicunt quod castrum cum burgo Veteris Sarum et dominicus burgus
domini Regis pertinent ad coronam cum advocatione cujusdam ecclesiæ quæ
modo vacat." _Hundred Rolls_, Edward I., cited by Benson and Hatcher,
p. 802.

[651] Cited by Benson and Hatcher, p. 802.

[652] D. B., 66a, 1. The value T. R. E. is not, however, very
distinctly stated.

[653] "Dicunt Angligenses burgenses de Sciropesberie multum grave sibi
esse quod ipsi reddunt totum geldum sicut reddebant T. R. E. quamvis
castellum comitis occupaverit 51 masuras et aliæ 50 masuræ sunt wastæ."
D. B., i., 252.

[654] Some writers, such as Mr Kerslake and Mr C. S. Taylor, have
supposed Sceargate to mean Shrewsbury.

[655] Mandatum est vicecomiti Salopie quod veterem palum et veterem
bretaschiam de vetere fossato ville Salopie faciat habere probos
homines ville Salopie ad novum fossatum ejusdem ville, quod fieri
fecerant, efforciandum et emendendum. _Close Rolls_, 1231, p. 508. The
honest men of the city are also to have "palum et closturam" from the
king's wood of Lichewood "ad hirucones circa villam Salopie faciendas
ad ipsam villam claudendam." _Ibid._ _Hirucones_ are the same as
_heritones_ or _hericias_, a defence of stakes on the counterscarp of
the ditch.

[656] "In op. castelli de Salop^be in mota 5_l._" _Pipe Rolls_, 19
Henry II., p. 108.

[657] "Dampnum mote castri Salopp' ad valenciam 60 marcarum, sed non
recolligunt totum evenisse propter molendinum abbatis Salopp', quia
30 annis elapsis mota castri fuit fere deteriorata sicut nunc est."
_Hundred Rolls_, ii., 80. "Dicunt quod unus magnus turris ligneus
(_sic_) qui ædificatur in castro Salopp' corruit in terram tempore
domini Uriani de S. Petro tunc vicecomitis, et meremium ejus turris
tempore suo et temporibus aliorum vicecomitum postea ita consumitur et
destruitur quod nihil de illo remansit, in magnum damnum domini Regis
et deteriorationem eiusdem castri." _Ibid._, p. 105.

[658] _Pipe Rolls_, 11 Henry II., p. 89; 12 Henry II., p. 59; 14 Henry
II., p. 93; 15 Henry II., p. 108; 20 Henry II., p. 108.

[659] Payment to those who dig stone for the castle of Shrewsbury,
_Close Rolls_, i., 622b. This is in 1224. There is also a payment of
50_l._ for works at the castle in 1223. _Ibid._, 533b.

[660] _Hundred Rolls_, ii., 80. A _jarola_ or garuillum is a stockade;
apparently derived from a Gallic word for _oak_, and may thus
correspond to an oak paling. See Ducange.

[661] Owen and Blakeway's _History of Shrewsbury_, i., 450.

[662] _Chronicon de Melsa_, R. S. See Preface, p. lxxii.

[663] _Yorks Inquisitions_ (Yorks Rec. Ser.), i., 83.

[664] _Rot. Lit. Claus._, i., 474b.

[665] Poulson's _History of Holderness_, i., 457.

[666] D. B., i., 323b.

[667] Ethelwerd, anno 910.

[668] "Ipse Henricus tenet Cebbeseio. Ad hoc manerium pertinuit terra
de Stadford, in qua rex precepit fieri castellum, quod modo est
destructum." D. B., i., 249a.

[669] "Apud Estafort alteram [munitionem] locavit." _Ord. Vit._, p. 199.

[670] It should be said that Mr Eyton interprets the passage
differently, and takes it to mean that the castle was built on land
in the borough of Stafford belonging to the manor of Chebsey. But he
himself says that "the site of Stafford Castle, within the liberties,
though not within the borough of Stafford, would suggest a royal
foundation"; and he believes this castle (the one on the motte) to have
been the one garrisoned by Henry I. and made a residence by Henry II.
_Domesday Studies_, p. 21.

[671] _Salt. Arch. Soc. Trans._, vol. viii., "The Manor of Castre or
Stafford," by Mr Mazzinghi, a paper abounding in valuable information,
to which the present writer is greatly indebted.

[672] In the addenda to Mr Eyton's _Domesday of Staffordshire_ (p. 135)
the learned editor says there are two Stafford castles mentioned in
Domesday, in two different hundreds. We have carefully searched through
the whole Stafford account, and except at Burton and Tutbury, there is
no other castle mentioned in Staffordshire but this one at Chebsey.

[673] Dugdale conjectures that Robert was sheriff of Staffordshire. He
had large estates round the town of Stafford. Eyton, _Staffordshire_,
p. 61.

[674] Mazzinghi, _Salt Arch. Soc. Trans._, viii., 6; Eyton, _Domesday
Studies_, p. 20.

[675] _Monasticon_, vi., 223: "Ecclesiam S. Nicholai in castello de
Stafford."

[676] Ordericus, vii., 12. See also vii., 13, p. 220 (ed. Prévost).

[677] Mazzinghi, _Salt Arch. Soc. Trans._, viii., 22.

[678] In a charter to Stone Abbey, _Salt Collections_, vol. ii. That
the castle he speaks of was the one outside the town is proved by his
references to land "extra burgum."

[679] The _Pipe Roll_ contains several entries relating to this gaol
at Stafford. It is clear from several of the documents given by Mr
Mazzinghi that the king's gaol of Stafford and the king's gaol of the
castle of Stafford are equivalent expressions.

[680] _Pipe Rolls_, 2 John.

[681] _Close Rolls_, i., 69.

[682] _Constitutional History_, i., 272.

[683] Cited in _Salt Arch. Soc. Trans._, vi., pt. i., 258.

[684] _Patent Rolls_, 22 Edward iii., cited by Mazzinghi, p. 80.

[685] _Salt Arch. Soc. Trans._, viii., 122. It was undoubtedly at this
time that the oblong stone keep on the motte, which is described in an
escheat of Henry VIII.'s reign, was built.

[686] _Salt Arch. Coll._, viii., 14.

[687] Speed's _Theatre of Britain_; Leland, _Itin._, vii., 26.

[688] The Stafford escheat of Henry VIII.'s reign, which describes the
town, also makes no mention of any castle in the town. Mazzinghi, p.
105.

[689] _Salt Arch. Trans._, viii., 231. The mistake may possibly have
arisen from the fact that a fine castellated gateway, shown in W.
Smith's map (_Description of England_), stood on the south-west wall of
the town, close to the spot where Speed's map marks a Castle Hill.

[690] There must be some error in the first statement of the Stafford
revenue in Domesday, which says that the king and earl have 7_l._
between them, as it is contradicted by the later statement. D. B., i.,
246a and 247b, 2.

[691] There were 141 _mansiones_, T. R. E., "et modo totidem sunt
præter 5 quæ propter operationem castelli sunt wastæ." From a passage
in the _Domesday of Nottingham_ it would seem that a _mansio_ was a
group of houses.

[692] _Gervase of Canterbury_, i., 156, R. S.

[693] Peck's _Antiquarian Annals of Stamford_; he gives the charter, p.
17.

[694] Cited in Nevinson's "Notes on the History of Stamford," _Journ.
Brit. Arch. Ass._, xxxv.

[695] "T. R. E. dabat Stanford 15_l._; modo dat ad firmam 50_l._ De
omni consuetudine regis modo dat 28_l._"

[696] "Ibi habet Helgot castellum, et 2 carucas in dominio, et 4
servos, et 3 villanos, et 3 bordarios, et 1 Francigenam cum 3-1/2
carucis. Ibi ecclesia et presbyter. T. R. E. valebat 18 solidos; modo
25 solidos. Wastam invenit." D. B., i., 258b. There are some fragments
of Norman work in the church, which is chiefly Early English, doubtless
of the same date as the mural tower of the castle.

[697] Stapleton's Introduction to _Rot. Scac. Normanniæ_, vol. ii.

[698] It used to be supposed that herring-bone work was a Saxon sign,
and this furnished an additional claim to the Saxon origin of this
castle; but it is now known that herring-bone work only occurs in the
later Saxon work, and is far more common in Norman. See _note_, p. 136.

[699] See _ante_, p. 34.

[700] Ordericus, xi., ch. iii.

[701] There are three entries for the works of the _turris_ at Tickhill
in the _Pipe Rolls_ of 1178 and 1179, amounting to £123, 12s. 5d.

[702] _Pipe Roll_, 31 Henry I., 33, 36. Expenses for work at the
wall of the castle are mentioned. Ordericus says that Robert Belesme
fortified the castle of Blythe at the time of his rebellion in 1101,
but he also says that it had belonged to Roger de Busli. _Hist. Ecc._,
iv., 33; xi., 3.

[703] Vicar's _Parliamentary Chronicle_, quoted by Hunter, _South
Yorks_, ii., 235.

[704] D. B., i., 319a.

[705] _A.-S. C._ in _anno_.

[706] D. B., i., 76.

[707] _M. A._, iv., 630.

[708] Leland is responsible for this last statement.

[709] D. B., i., 108b.

[710] "Egressus Lundoniæ rex _dies aliquot_ in propinquo loco Bercingio
morabatur, dum firmamenta quædam in urbe contra mobilitatem ingentis et
feri populi perficerentur." P. 165. Ordericus is quoting from William
of Poitiers. There was formerly a Roman camp at Barking, and the motte
which William hastily threw up on its rampart to defend his sojourn
still remains. See _Victoria History of Essex_.

[711] Mr Harold Sands suggests to me that the first fortification may
simply have been a bank and palisade across the angle of the Roman
wall, with perhaps a wooden keep, and that the great fire in London in
1077 determined William to build a stone keep.

[712] Hearne's _Textus Roffensis_, 212. "Idem Gundulfus, ex precepto
Regis Willielmi Magni, præesset operi magnæ turris Londoniæ."

[713] The building of stone keeps was generally spread over several
years, as we learn from the _Pipe Rolls_. Richard I. built his
celebrated keep of Chateau Gaillard in one year, but he himself
regarded this as an architectural feat. "Estne bella, filia mea de uno
anno," he said in delight.

[714] _A.-S. C._ in _anno_.

[715] Round's _History of Colchester_, ch. iv.

[716] The keep of Norwich Castle measures 100 × 95 feet; Middleham, 100
× 80; Dover, 95 × 90. These are the largest existing keeps in England,
next to the Tower and Colchester. The destroyed keep of Duffield
measured 99 × 93 feet; that of Bristol is believed to have been 110 ×
95.

[717] The reader will find little help for the structural history of
the Tower in most of the works which call themselves Histories of the
Tower of London. The plan of these works generally is to skim over the
structural history as quickly as possible, perhaps with the help of a
few passages from Clark, and to get on to the history of the prisoners
in the Tower. For the description in the text, the writer is greatly
indebted to Mr Harold Sands, F.S.A., who has made a careful study of
the Tower, and whose monograph upon it, it is hoped, will shortly
appear.

[718] _Ante_, p. 89.

[719] Many of the larger keeps contain rooms quite spacious enough to
have served as banqueting halls, and it is a point of some difficulty
whether they were built to be used as such. But as late as the 14th
century, Piers Ploughman rebukes the new custom which was growing up of
the noble and his family taking their meals in private, and leaving the
hall to their retainers. Every castle seems to have had a hall in the
bailey.

[720] Mr Sands says the main floors are not of too great a span to
carry any ordinary weight.

[721] The keep of Pevensey Castle, the basement of which has been
recently uncovered, has no less than four apsidal projections, one of
which rests on the solid base of a Roman mural tower. But this keep is
quite an exceptional building. See _Excavations at Pevensey_, Second
Report, by H. Sands.

[722] Mr Sands has conjectured that the third floor may be an addition,
and that the second storey was originally open up to the roof and
not communicating with the mural passage except by stairs. This was
actually the case at Bamborough keep, and at Newcastle and Rochester
the mural gallery opens into the upper part of the second storey by
inner windows.

[723] Until the end of the 12th century the roofs of keeps were gabled
and not flat, but probably there was usually a parapet walk for
sentinels or archers.

[724] Parts of these walls, running N. and S. have been found very near
the E. side of the Tower. No trace of the Roman wall has been found S.
of the Tower, but in Lower Thames Street lines have been found which,
if produced, would lead straight to the S. wall of the inner bailey.
Communicated by Mr Harold Sands.

[725] I have to thank Mr Harold Sands for kindly revising this account
of the Tower.

[726] "Ibi habet comes unum castrum et mercatum, reddentes 101s." D.
B., i., 122.

[727] It must be remembered that round arches, in castle architecture,
are by no means a certain sign of date. Of course the first castle on
this motte must have been of wood.

[728] _Ord. Vit._, ii., 222 (Prévost).

[729] "Henricus de Ferrers habet castellum de Toteberie. In burgo circa
castellum sunt 42 homines de mercato suo tantum viventes." D. B., i.,
248b.

[730] Shaw's _History of Staffordshire_, i., 49.

[731] Quoted in _Beauties of England and Wales_, Staffordshire, p. 1129.

[732] _Diceto_, i., 384. The castle was then besieged on Henry's behalf
by the vassal prince of South Wales, the Lord Rhys.

[733] The foundation charter is in _Mon. Ang._, iii., 393.

[734] _A.-S. C._

[735] William of Poitiers calls it an _oppidum_, p. 141.

[736] Hedges, _History of Wallingford_.

[737] "The Towne of Portsmuth is murid from the Est Tower a forowgh
lenght with a Mudde Waulle armid with Tymbre." _Itin._, iii., 113.

[738] "In burgo de Walingeford habuit Rex Edwardus 8 virgatas terræ;
et in his erant 276 hagæ reddentes 11 libras de gablo.... Pro castello
sunt 8 destructæ." D. B., i., 56. If we divide these 276 _haughs_
by the 114 acres enclosed by the town rampart, we get an average of
about 1 rood 26 perches for each haugh; multiply this by 8 (the number
destroyed for the castle) and we get an area of 3 acres, which is about
the average area of an early Norman castle.

[739] Hedges, _History of Wallingford_, i., 139.

[740] Camden speaks of the motte as being in the middle of the castle,
but this is a mistake.

[741] Such is the account in Hedges' _History of Wallingford_, p. 139,
but it sounds odd. It is to be inferred from the same source that the
fragment of a round building which stands on the top of the motte must
be modern; it is thick enough to be ancient.

[742] _Close Rolls_, i., anno 1223.

[743] D. B., i., 56.

[744] "Abbas de Couentreu habet 36 masuras, et 4 sunt wastæ propter
situm castelli." D. B., i., 238a.

[745] "Hæ masuræ pertinent ad terras quas ipsi barones tenent extra
burgum, et ibi appreciatæ sunt." D. B., i., 238.

[746] Maitland, _Domesday Book and Beyond_, p. 189.

[747] Ordericus, p. 184. "Rex _itaque_ castellum apud Guarevicum
condidit, et Henrico Rogerii de Bello Monte filio ad servandum
tradidit." Mr Freeman remarks that no authentic records connect Thurkil
of Warwick with Warwick Castle. _N. C._, iv., 781.

[748] _N. C._, iv., 190.

[749] In operatione unius domus in mota de Warwick et unius bretaschie
5_l._ 7_s._ 11_d._ _Pipe Rolls_, 20 Henry II. As _domus_ is a word very
commonly used for a keep, it is probable this expenditure refers to a
wooden keep.

[750] From information received from Mr Harold Sands. There appears to
be no foundation whatever for the curious ground plan given by Parker.

[751] See _ante_, p. 42.

[752] "Willelmus comes fecit illud castellum in wasta terra quæ vocatur
Mereston." D. B., i., 183.

[753] _Mon. Ang._, vi., 349.

[754] This keep rests on a broad extension of the earthen rampart,
similar to what is still to be seen in the mottes of Devizes,
Burton-in-Lonsdale, and William Hill, Middleham.

[755] Ordericus says: "Intra moenia Guentæ, opibus et munimine nobilis
urbis et mari contiguæ, validam arcem construxit, ibique Willelmum
Osberni filium in exercitu suo precipuum reliquit." II., 166. The
_intra moenia_ is not to be taken literally, any more than the _mari
contiguæ_. It is strange that Mr Freeman should have mistaken Guenta
for Norwich, since under 1067 Ordericus translates the Winchester of
the A.-S. C. by Guenta.

[756] "De isto manerio testatur comitatus quod injuste accepit [abbas]
pro excambio domus regis, quia domus erat regis." D. B., i., 43a, 1.

[757] _Ibid._, i., 43a, 2.

[758] "Sicut rex Willielmus pater meus ei dedit in excambium pro terra
illa in qua ædificavit aulam suam in urbe Winton." _Mon. Ang._, ii.,
444.

[759] "Pars erat in dominio et pars de dominio abbatis; hoc totum est
post occupatum in domo regis." P. 534. This passage throws light on the
fraud of the abbot of Hyde, referred to above.

[760] "Extra portam de Vuest ... ibi juxta fuit quidam vicus; fuit
diffactus quando rex fecit facere suum fossatum." P. 535.

[761] _Arch. Inst._, Winchester volume, p. 51.

[762] It should also be said that the word _domus_ is frequently used
for a keep in chronicles and ancient documents of the 11th and 12th
centuries.

[763] The line of the more ancient roof gable can be traced in the
north wall, and there is a vestige of a Norman doorway in the east wall.

[764] _History of Winchester_, ii., 210.

[765] Henry of Blois, Bishop of Winchester and brother of King Stephen,
pulled down the royal palace close to the cathedral, which presumably
was the old Saxon palace, and used the materials to build Wolvesey
Castle. See Malmesbury, "De Vitis Sex Episcoporum," _Anglia Sacra_,
ii., 421. He could hardly have dared to do this if the palace had still
been used by the Norman kings.

[766] _History of Winchester_, ii., 210. See Fig. 37.

[767] _Ibid._, p. 195. It is difficult, now that the area has been
levelled, to say exactly where this motte stood. Woodward says that the
keep stood in the N.E. corner; but he probably alludes to a mural tower
whose foundations can still be seen, near the County Hall. _History of
Hampshire_, i., 295-304.

[768] Turner, _History of Domestic Architecture_. He cites from the
_Liberate Roll_, 35 Henry II., an order for the repair of the ditch
between the great tower and the bailey.

[769] "Radulfus filius Seifrid tenet de rege Clivor. Heraldus comes
tenuit. Tunc se defendebat pro 5 hidis, modo pro 4-1/2 hidis, et
castellum de Windesores est in dimidia hida." D. B., i., 62b. The
_Abingdon History_ also mentions the foundation of Windsor Castle and
gives some interesting details about castle guard. "Tunc Walingaforde
et Oxenforde et Wildesore, cæterisque locis, castella pro regno
servando compacta. Unde huic abbatiæ militum excubias apud ipsum
Wildesore oppidum habendas regis imperio jussum." II., 3, R. S.

[770] _Leland_, iv., 1, 37. See also Tighe's _Annals of Windsor_, pp.
1-6. Until recently there was a farmhouse surrounded by a moat at Old
Windsor, which was _believed_ to mark the site of Edward's _regia
domus_.

[771] Edward's grant of Windsor to Westminster is in _Cod. Dip._, iv.,
227. Domesday does not mention the rights of the church, but says the
manor of Windsor was held of the crown T. R. E. and T. R. W. Camden
gives William's charter of exchange with the convent of Westminster.
_Britannia_, i., 151.

[772] This is stated in the charter given by Camden.

[773] In 1 virgata terræ quam Willelmus fil. Walteri habet in escambio
pro terra sua quæ capta est ad burgum. P. 721.

[774] The _Red Book of the Exchequer_, which contains an abstract of
the missing _Pipe Roll_ of 1 Henry II., has an entry of 12_s._ paid to
Richard de Clifwar for the exchange of his land, and regular payments
are made later. There was another enlargement of the bailey in Henry
III.'s reign, but the second bailey was then existing. See _Close
Rolls_, i., 531b.

[775] "In operatione muri circa castellum 11_l._ 10_s._ 4_d._ Summa
denariorum quos idem Ricardus [de Luci] misit in operatione predicta de
ballia 128_l._ 9_s._" _Pipe Roll_, 20 Henry II., p. 116.

[776] Tighe's _Annals of Windsor_, p. 21.

[777] There is a singular entry in the _Pipe Roll_ of 7 Richard I.,
"pro fossato prosternando quod fuit inter motam et domos regis,"
clearly the ditch between the motte and the bailey. Mr Hope informs
me that this can only refer to the northern part of the ditch, as the
eastern portion was only filled up in 1824. Mr Hope thinks that the
castle area has always included the lower bailey. I regret that Mr
Hope's History of Windsor Castle did not appear in time to be used in
this work.

[778] _Foedera_, vol. i.

[779] _Pipe Rolls_, 30 Henry II.

[780] D. B., i., 62b, 2; 56b, 2.

[781] Roger of Wendover, in _anno_.

[782] Walter and Cradock's _History of Wisbeach_, pp. 270-278.

[783] Morris' _Troubles of our Catholic Forefathers_, p. 223. This keep
was one built by Bishop Morton in 1471.

[784] Birch's _Cartularium_, ii., 222.

[785] Ursus erat vicecomes Wigorniæ a rege constitutus, qui in ipsis
poene faucis monachorum castellum construxit, adeo ut fossatum
coemiterii partem decideret. _Gesta Pontif._, p. 253.

[786] "Castrum Wigorniæ nobis redditum est, tanquam jus noster, usquam
motam turris." _Annales de Wigornia_, R. S., p. 407. "Rex Johanni
Marescallo salutem: Mandamus vobis quod sine dilatione faciatis habere
venerabili patri nostro domino Wigorniensi episcopo ballium castri
nostri Wigorniæ, quod est jus ecclesiæ suæ; retenta ad opus nostrum
mota ejusdem castri." _Patent Rolls_, 1 Henry III., p. 46.

[787] _Annales de Wigornia_, p. 375.

[788] "In reparatione turris Wigorniæ 8_l._" _Red Book of Exchequer_,
ii., 656.

[789] "Precipimus tibi quod per visum liberorum et legalium hominum
facias parari portam castri Wigorniæ, quæ nunc est lignea, lapideam, et
bonam et pulchram." _Rot. de Liberate_, p. 93, 1204.

[790] Green's _History of Worcester_, i., 19.

[791] Allies' _Antiquities of Worcestershire_, p. 15. His words
strictly apply to "the lofty mound called the keep, with its ditches,
etc.," but probably the whole area was not more than 4 acres.

[792] See the documents cited by Mr Round in his _Geoffrey de
Mandeville_, Appendix O, and the _Pipe Rolls_ of 1173. "In reparatione
Mote et Gaiole de Wirecestra, £35, 13s. 8d."

[793] _Gentleman's Magazine_, i., 36, 1834. See Haverfield,
"Romano-British Worcester," _Victoria County History of
Worcestershire_, vol. i.

[794] D. B., i., 172.

[795] It is needless to remark that _baile_ is the Norman word for an
enclosure or courtyard; Low Latin _ballia_; sometimes believed to be
derived from _baculus_, a stick.

[796] Ordericus, ii., 188 (edition Prévost).

[797] _Norman Conquest_, iv., 270. Mr Freeman has worked out the course
of events connected with the building and destruction of the castles
with his usual lucidity. But he never grasped the real significance of
mottes, though he emphatically maintained that the native English did
not build castles.

[798] "Ethelstanus Castrum quod olim Dani in Eboraco obfirmaverant ad
solum diruit, ne esset quo se tutari perfidia posset." _Gesta Regum_,
ii., 134.

[799] Widdrington, _Analecta Eboracensia_, p. 120. It was this suburb
which Alan, Earl of Richmond gave to the Abbey of St Mary at York,
which he had founded. "Ecclesiam sancti Olavii in quâ capud abbatiæ
in honorem sanctæ Mariæ melius constitutum est, et _burgum in quo
ecclesia sita est_." _Mon. Ang._, iii., 547. For the addition of new
boroughs to old ones see _ante_, p. 174, under Norwich. Although
Athelstan destroyed the fortifications of this borough, they were
evidently renewed when the Danish earls took up their residence there,
for when Earl Alan persuaded the monks from Whitby to settle there one
inducement which he offered was the fortification of the site, "loci
munitionem." _Mon. Ang._, iii., 545.

[800] In Eboraco civitate T. R. E. præter scyram archiepiscopi fuerunt
6 scyræ; una ex his est wasta in castellis. D. B., i., 298.

[801] _Notes on Clifford's Tower_, by George Benson and H. Platnauer,
published by the York Philosophical Society.

[802] "Thone castel tobræcon and towurpan." _A.-S. C._ See Freeman, _N.
C._, iv., 270.

[803] "In operatione turris de Euerwick, 15_l._ 7_s._ 3_d._" _Pipe
Roll_, 19 Henry II., vol. xix., 2. We assume that William's second keep
lasted till Henry II.'s reign.

[804] _Benedict of Peterborough_, ii., 107.

[805] "In operatione castri 28_l._ 13_s._ 9_d._" _Pipe Roll_, 3 Richard
I. Under the year 1193, after relating the tragedy of the Jews at York
Castle, Hoveden says: "Deinde idem cancellarius [William de Longchamp]
tradidit Osberto de Lunchamp, fratri suo, comitatum Eboracensem in
custodia, et precepit firmari castellum in veteri castellario quod
Rex Willelmus Rufus ibi construxerat." III., 34, R. S. The expression
_vetus castellarium_ would lead us to think of the Old Baile, which
certainly had this name from an early period; and Hoveden, being a
Yorkshireman as well as a very accurate writer, was probably aware of
the difference between the two castles. But if he meant the Old Baile,
then both the castles were restored at about the same time. "Rufus"
must be a slip, unless there was some rebuilding in Rufus' reign of
which we do not know.

[806] Messrs Benson and Platnauer are of the former opinion. "The
existence of a second layer of timber seems to show that the
fortification destroyed was rebuilt in wood." _Notes on Clifford's
Tower_, p. 2.

[807] "Pro mairemio castri Ebor. prostrato per ventum colligendo,
2_s._" _Pipe Roll_, 19 Henry III. It is, of course, a conjecture that
this accident happened to the keep; but the keep would be the part most
exposed to the wind, and the _scattering_ of the timber, so that it had
to be collected, is just what would happen if a timber structure were
blown off a motte.

[808] As the writer was the first to publish this statement, it will
be well to give the evidence on which it rests. The keep of York is
clearly Early English in style, and of an early phase of the style. It
is, however, evident to every one who has carefully compared our dated
keeps, that castle architecture always lags behind church architecture
in style-development, and must be judged by different standards. We
should therefore be prepared to find this and most other keeps to be
of later date than their architecture would suggest. Moreover, the
expenditure entered to York Castle in the reigns of Henry II., Richard
I., and John, is quite insufficient to cover the cost of a stone keep.
The _Pipe Rolls_ of Henry III.'s reign decide the matter, as they show
the sums which he expended annually on this castle. It is true they
never mention the _turris_, but always the _castrum_; we must also
admit that the _turris_ and _castrum_ are often distinguished in the
writs, even as late as Edward III.'s reign. (_Close Rolls_, 1334.) On
the other hand extensive acquaintance with the _Pipe Rolls_ proves that
though the mediæval scribe may have an occasional fit of accuracy, he
is generally very loose in his use of words, and his distinctions must
never be pressed. Take, for instance, the case of Orford, where the
word used in the _Pipe Rolls_ is always _castellum_, but it certainly
refers to the keep, as there are no other buildings at Orford. Other
instances might be given in which the word _castellum_ clearly applies
to the keep. It should be mentioned that in 1204 John gave an order
for stone for the castle (_Close Rolls_, i., 4b), but the amounts on
the bill for it in the _Pipe Rolls_ show that it was not used for any
extensive building operations.

[809] "Mandatum est Galterio de Cumpton forestario de Gauteris quod
ad pontem et domos castri Eboraci et breccas palicii ejusdem castri
reparandos et emendandos Vicecomitem Eboraci mæremium habere faciat in
foresta de Gauteris per visum, etc." _Close Rolls_, ii., 61b.

[810] Order to expend up to 6 marks in repairing the wooden peel about
the keep of York Castle, which peel is now fallen down. _Cal. of Close
Rolls,_ 17 Edward II., 25.

[811] _Cal. of Close Rolls_, 1313-1318, 262. _Mota_ is wrongly
translated _moat_.

[812] See Mr Cooper's _York: The Story of its Walls and Castles_.
During Messrs Benson and Platnauer's excavations, a prehistoric
crouching burial was found in the ground below the motte, 4 feet 6
inches under the present level. This raises the question whether
William utilised an existing prehistoric barrow for the nucleus of his
motte.

[813] D. B., i., 298a.

[814] _York: The Story of its Walls and Castles_, by T. P. Cooper, p.
222.

[815] See the passage from Hoveden already quoted, _ante_, p. 245.

[816] Drake's _Eboracum_, App. xliv.

[817] See Mr Cooper's _York: The Story of its Walls and Castles_, which
contains a mass of new material from documentary sources, and sheds
quite unexpected light on the history of the York fortifications. I am
indebted to Mr Cooper's courtesy for some of the extracts cited above
relating to York Castle.

[818] Cooper's _York_, chapters ii. and iv. 100_l_. was spent by the
sheriff in fortifying the walls of York in the sixth year of Henry III.
After this there are repeated grants for murage in the same and the
following reign. There are some Early English buttresses in the walls,
but the majority are later. No part of the walls contains Norman work.

[819] The details of this evidence, which consist mainly in (1) a
structural difference in the extended rampart; (2) a subsidence in the
ground marking the old line of the city ditch, will be found in Mr
Cooper's work, p. 224.

[820] "Locum in Eboraco qui dicitur Vetus Ballium, primo spissis et
longis 18 pedum tabulis, secundo lapideo muro fortiter includebat." T.
Stubbs, in Raine's _Historians of the Church of York_, ii., 417, R. S.

[821] "The plotte of this castelle is now caullid the Olde Baile, and
the area and diches of it do manifestley appere." _Itin._, i., 60.

[822] See the plan in Mr Cooper's _York_, p. 217.

[823] "In the Wales of the Laws, the social system is tribal." Owen
Edwards, _Wales_, p. 39.

[824] Vinogradoff, _Growth of the Manor_, pp. 15-16.

[825] Pennant's _Tour in Wales_, Rhys' edition, ii., 234.

[826] _Ancient Laws and Institutes of Wales_, pp. 238, 94. The MS. of
the _Leges Wallicæ_ is not earlier than the 13th century. The other
editions of the Laws are even later. See Wade Evans, _Welsh Mediæval
Law_, for the most recent criticism of the Laws of Howel Dda.

[827] The _Leges Wallicæ_ say: "Villani regis debent facere novem domos
ad opus regis; scilicet, aulam, cameram, coquinam, penu (capellam),
stabulum, kynorty (stabulum canum), horreum, odyn (siccarium) et
latrinam." P. 791.

[828] The word Din or Dinas, so often used for a fort in Wales, is
cognate with the German _Zaun_, Anglo-Saxon _tun_, and means a fenced
place. Neither it nor the Irish form dun have any connection with the
Anglo-Saxon _dun_, a hill. See J. E. Lloyd, _Welsh Place-names_, "Y
Cymmrodor," xi., 24.

[829] It is doubtful whether Deheubarth ever included the small
independent states of Gwent, Brecknock, and Glamorgan.

[830] "Wales and the Coming of the Normans," _Cymmrodorion Trans._,
1899.

[831] There is an earthwork near Portskewet, a semicircular cliff camp
with three ramparts and two ditches. It is scarcely likely that this
can be Harold's work, as Roman bricks are said to have been found
there. Willet's _Monmouthshire_, p. 244. Athelstan had made the Wye the
frontier of Wales. _Malmesbury_, ii., 134.

[832] See _A.-S. C._, anno 1097, and compare the entry for 1096 with
the account in the _Brut_ for 1093, which shows that the Norman castles
had been restored, after being for the most part demolished by the
Welsh.

[833] The _Brut y Tywysogion_, or _Story of the Princes_, exists in
no MS. older than the 14th century. It and the _Annales Cambriæ_ have
been disgracefully edited for the _Rolls_ Series, and the topographical
student will find no help from these editions. See Mr Phillimore's
criticism of them, in _Y Cymmrodor,_ vol. xi. The Aberpergwm MS. of
the _Brut_, known also as the _Gwentian Chronicle_, has been printed
in the _Archæologia Cambrensis_ for 1864; it contains a great deal of
additional information, but as Mr Phillimore observes, so much of it is
forgery that none of it can be trusted when unsupported.

[834] The barbarity on both sides was frightful, but in the case of
the Welsh, it was often their own countrymen, and even near relations,
who were the victims. And so little patriotism existed then in Wales
that the Normans could always find allies amongst some of the Welsh
chieftains. Patriotism, however, is a virtue of more recent growth than
the 11th century.

[835] There is, however, no contemporary evidence for the existence of
the Marcher lordships before the end of the 12th century. See Duckett
"On the Marches of Wales," _Arch. Camb._, 1881.

[836] The districts of Cyfeiliog and Arwystli, in the centre of Wales,
were also reckoned in Gwynedd.

[837] "Wales and the Coming of the Normans," _Cymmrodorion Trans._,
1899.

[838] In the descriptions of castles in this chapter, those which have
not been specially visited for this work are marked with an asterisk.
Those which have been visited by others than the writer are marked with
initials: D. H. M. being Mr D. H. Montgomerie, F.S.A.; B. T. S., Mr
Basil T. Stallybrass; and H. W., the Rev. Herbert White, M.A. This plan
will be followed in the three succeeding chapters.

[839] "Hugo comes tenet de rege Roelent (Rhuddlan). Ibi T. R. E.
jacebat Englefield, et tota erat wasta. Edwinus comes tenebat. Quando
Hugo comes recipit similiter erat wasta. Modo habet in dominio
medietatem castelli quod Roelent vocatur, et caput est hujus terræ....
Robertus de Roelent tenet de Hugone comite medietatem ejusdem castelli
et burgi, in quo habet ipse Robertus 10 burgenses et medietatem
ecclesiæ. Ibi est novus burgus et in eo 10 burgenses.... In ipso
manerio est factum noviter castellum similiter Roeland appellatum." D.
B., i., 269a, 1.

[840] Ayloffe's _Rotuli Walliæ_, p. 75. "De providendo indempnitati
magistri Ricardi Bernard, Personæ Ecclesiæ de Rothelan', in
recompensionem terræ suæ occupatæ ad placeam castri de Rothelan'
elargandam."

[841] Tut or Toot Hill means "look-out" hill; the name is not
unfrequently given to abandoned mottes. The word is still used locally.
_Cf._ Christison, _Early Fortifications in Scotland_, p. 16.

[842] Such presentations of abandoned castle sites, and of old wooden
castles, to the church, were not uncommon. We have seen how the
site of Montacute Castle was given to the Cluniac monks (_ante_, p.
170). Thicket Priory, in Yorkshire, occupied the site of the castle
of Wheldrake; and William de Albini gave the site and materials of
the old castle of Buckenham, in Norfolk, to the new castle which he
founded there. The materials, but not the site, of the wooden castle of
Montferrand were given in Stephen's reign to Meaux Abbey, and served to
build some of the monastic offices. _Chron. de Melsa_, i., 106.

[843] "Fines suos dilatavit, et in monte Dagannoth, qui mari contiguus
est, fortissimum castellum condidit." _Ordericus_, iii., 284 (edition
Prévost). The verb _condere_ is never used except for a new foundation.

[844] The _Brut_ says that in the year 823 the Saxons destroyed the
_Castle_ of Deganwy. This is one of the only two instances in which
the word _castell_ is used in this Welsh chronicle before the coming
of the Normans. As the MS. is not earlier than the 14th century it
would be idle to claim this as a proof of the existence of a castle at
this period. _Castell_, in Welsh, is believed to have come straight
from the Latin, and was applied to any kind of fortress. Lloyd, _Welsh
Place-names_, "Y Cymmrodor," xi., 28.

[845] The "new castle of Aberconwy" mentioned by the _Brut_ in 1211,
undoubtedly means this new stone castle built by the earl at Deganwy,
as the castle of Conway did not then exist.

[846] See Pennant, ii., 151; and _Arch. Camb._, 1891, p. 321.

[847] _Brut of Tywysogion_, 1145.

[848] Published with a Latin translation in _Arch. Camb._, 1866. "He
built castles in various places, after the manner of the French, in
order that he might better hold the country."

[849] The _Brut_ also mentions the castle of Aberlleinog, and says it
was built in 1096; _rebuilt_ would have been more correct, as the "Life
of Griffith ap Cynan" shows that it was built by the Earl of Chester,
and burnt by Griffith, before the expedition of 1096 (really 1098),
when Hugh, Earl of Shrewsbury, met with his death on the shore near
this castle, from an arrow shot by King Magnus Barefoot, who came to
the help of the Welsh.

[850] Mr Hartshorne in his paper on Carnarvon Castle (_Arch. Journ._,
vii.) cites a document stating that a wall 18 perches long had been
begun _round the moat_ [possibly _motam_; original not given]. He also
cites from the _Pipe Rolls_ an item for wages to _carriers of earth dug
out of the castle_.

[851] This ruined wall runs in a straight line through the wood on the
ridge to the east of the town; at one place it turns at right angles;
at the back of the golf pavilion is a portion still erect, showing that
it was a dry built wall of very ordinary character.

[852] Roman masonry has been exposed in the bank of the station.

[853] _Life of Griffith ap Cynan_; Brut, 1111.

[854] _Arch. Camb._, iv., series 296 and 911.

[855] The _Anglo-Saxon Chronicle_ dates this expedition in 1114,
and says that Henry caused castles to be built in Wales. The _Brut_
mentions the large tribute, 1111.

[856] _Brut_, 1149. Madoc ap Meredith, with the assistance of Ranulf,
Earl of Chester, prepared to rise against Owen Gwynedd, son of Griffith
ap Cynan.

[857] D. B., i., 255a. Professor Lloyd says, "Maelor Saesneg, Cydewain,
Ceri, and Arwystli came under Norman authority, and paid renders of
money or kine in token of subjection." "Wales and the Coming of the
Normans," _Cymmrodor. Trans._, 1899.

[858] _Ibid._

[859] See page 130.

[860] _Brut_, under 1107. The castle is called Dingeraint by this
chronicler.

[861] "Ipse comes construxit castrum Muntgumeri vocatum." D. B., i.,
254.

[862] _Montgomery Collections_, x., 56.

[863] _Close Rolls_, i., 558b.

[864] "Firmiter precipimus omnibus illis qui motas habent in valle de
Muntgumeri quod sine dilatione motas suas bonis bretaschiis firmari
faciant ad securitatem et defensionem suam et partium illarum." _Close
Rolls_, ii., 42.

[865] Mr Davies Pryce has suggested that the Hen Domen, a very perfect
motte and bailey within a mile of the present castle of Montgomery was
the original castle of Montgomery, and that the one built by Henry III.
was on a new site. This of course is quite possible, but I do not see
that there is sufficient evidence for it. See _Eng. Hist. Rev._, xx.,
709.

[866] _Brut y Tywysogion._

[867] _Itin._, vii., 16.

[868] _Pipe Rolls_, 1158-1164. It should be noted that the _Brut_ does
not claim the battle of Crogen as a Welsh victory.

[869] Lyttleton's _History of Henry II._

[870] Pennant thought he saw vestiges of a castle "in the foundations
of a wall opposite the ruins" [of the abbey]; but his accuracy is not
unimpeachable.

[871] _Pipe Rolls_, 1211-1213. "For the money expended in rescuing the
castles of Haliwell and Madrael, £100."

[872] _Itin._, p. 67. Toulmin Smith's edition of Welsh portion.

[873] D. B., i., 255a.

[874] Life of Griffith.

[875] _Pipe Roll_, 1159-1160. £4, 3s. 4d. paid to Roger de Powys "ad
custodiam castelli de Dernio"; "In munitione turris de Dermant £6, 4s.
0d." It cannot be doubted that these two names mean the same place.

[876] _Arch. Camb._, iv., 1887.

[877] At the time of the Survey the manor of Gresford (Gretford) was
divided between Hugh, Osbern, and Rainald. Osbern had 6-1/2 hides and a
mill grinding the corn of _his court_ (curiæ suæ). This probably is a
reference to this castle. D. B., i., 268. It was waste T. R. E. but is
now worth £3, 5s. 0d.

[878] "On the Town of Holt," by A. N. Palmer, _Arch. Camb._, 1907.

[879] _Beauties of England and Wales, North Wales_, p. 589. I am glad
to find that Mr Palmer, in the new edition of his _Ancient Tenures of
Land in the Marches of Wales_, confirms the identifications which I
have made of these two last castles, pp. 108, 116, 118.

[880] _Arch. Camb._, 5th ser., iv., 352. Camden's statement that this
castle was founded in Edward I.'s reign shows that he was unacquainted
with the _Pipe Rolls_.

[881] _Pipe Rolls_, 1164-1165, and 1167-1168.

[882] _Pipe Rolls_, 1212-1213.

[883] "Sur l'ewe de Keyroc," _History of Fulk Fitz Warine_, edited by
T. Wright for Warton Club.

[884] _Victoria County History of Lancashire_, i., 369.

[885] _England under the Normans and Angevins._

[886] "Ad recutienda castella de Haliwell et Madrael £100." _Pipe
Rolls_, 1212-1213.

[887] Wade Evans, _Welsh Mediæval Law_, vol. xii.

[888] It has in fact every appearance of a Roman camp.

[889] _Brut_, 1211.

[890] The castle of Hawarden, which is only about 2-1/4 miles from
that of Euloe, is not mentioned in any records before 1215; but it is
believed to have been a castle of the Norman lords of Mold. It also is
on a motte.

[891] I am indebted for this identification to the kindness of Mr A. N.
Palmer of Wrexham.

[892] D. B., i., 254. The manor is called Gal. It had been waste T. R.
E., but was now worth 40s.

[893] _Pipe Roll_ (unpublished), 1212-1213.

[894] Whereas there is no rock in the ditch of the neighbouring
motte of Tomen y Rhodwydd. Pennant (and others following him) most
inaccurately describe Tomen y Rhodwydd as _two_ artificial mounts,
whereas there is only one, with the usual embanked court. See Appendix
K.

[895] "The Maer dref [which Vardra represents] may be described as
the home farm of the chieftain." Rhys and Brynmor Jones, _The Welsh
People_, p. 401.

[896] Ordericus, ii., 218, 219 (edition Prévost).

[897] _Brut y Tywysogion_, 1091.

[898] _Brut_, 1071. "The French ravage Ceredigion (Cardigan) and
Dyfed"; 1072, "The French devastated Ceredigion a second time."

[899] _A.-S. C._, 1081. "This year the king led an army into Wales,
and there he set free many hundred persons"--doubtless, as Mr Freeman
remarks, captives taken previously by the Welsh. The _Brut_ treats this
expedition as merely a pilgrimage to St David's!

[900] "Then the French came into Dyfed and Ceredigion, _which they have
still retained_, and fortified the castles, and seized upon all the
land of the Britons." _Brut_, 1091 = 1093.

[901] Powell's _History of Wales_ professes to be founded on that of
Caradoc, a Welsh monk of the 12th century; but it is impossible to say
how much of it is Caradoc, and how much Powell, or Wynne, his augmentor.

[902] _Brut_, 1107.

[903] "In the Brut, Ystrad Towy does not only mean the vale of Towy,
but a very large district, embracing most of Carmarthenshire and
part of Glamorganshire." _Welsh Historical Documents_, by Egerton
Phillimore, in _Cymmrodor_, vol. xi.

[904] _Brut_, 1092.

[905] Lloyd, "Wales and the Coming of the Normans," _Cymmrodor.
Trans._, 1899: refers to Marchegay, _Chartes du Prieurie de Monmouth_.

[906] _Brut_, 1143.

[907] The date given is 1080, but as the dates in the Brut at this
period are uniformly two years too early, we alter them accordingly
throughout this chapter.

[908] Now more often called the Aberpergwm Brut, from the place where
the MS. is preserved.

[909] See Freeman, _Norman Conquest_, v., 820; William Rufus, ii., 79;
and Prof. Tout, in Y Cymmerodor, ix., 208. For this reason we do not
use the list of castles given in this chronicle, but confine ourselves
to those mentioned in the more trustworthy _Brut y Tywysogion_.

[910] The same MS. says, under the year 1099, "Harry Beaumont came to
Gower, against the sons of Caradog ap Jestin, and won many of their
lands, and built the castle of Abertawy (Swansea) and the castle of
Aberllychor (Loughor), and the castle of Llanrhidian (Weobley), and the
castle of Penrhys (Penrice), and established himself there, and brought
Saxons from Somerset there, where they obtained lands; and the greatest
usurpation of all the Frenchmen was his in Gower."

[911] "Primus hoc castrum Arnulphus de Mongumeri sub Anglorum rege
Henrico primo ex virgis et cespite, tenue satis et exile construxit."
_Itin. Cambriæ_, R. S., 89.

[912] Quoted from Duchesne in _Mon. Ang._, vol. vi.

[913] See Mr Cobbe's paper on Pembroke Castle in _Arch. Camb._,
1883, where reasons are given for thinking that the present ward was
originally, and even up to 1300, the whole castle.

[914] A motte-castle of earth and wood was certainly not regarded as "a
weak and slender defence" in the time of Giraldus.

[915] _Brut y Tywysogion_, 1095.

[916] Bridgeman's _Hist. of South Wales_, 17.

[917] _Arch. Camb._, 3rd ser., v., a paper on Newport Castle, in which
the writer says that there are _two_ mottes at Llanhyfer, the larger
one ditched round. The Ordnance Map only shows one.

[918] _Brut y Tywysogion_, 1146.

[919] _Patent Rolls of Henry III._, 255; _Foedera_, i., 161.

[920] _Brut y Tywysogion_, 1192.

[921] Bridgeman says that Narberth was given to Stephen Perrot by
Arnulf de Montgomeri, but gives no authority for this statement.

[922] _Brut_, 1171.

[923] _Ibid._, 1107. "Earl Gilbert built a castle at Dingeraint, where
Earl Roger had before founded a castle."

[924] The castle of Aberrheiddiol is probably the name of the present
castle of Aberystwyth when it was first built, as Lewis Morris says
that the river Rheiddiol formerly entered the sea near that point.
Quoted by Meyrick, _History of Cardigan_, p. 488.

[925] _Brut_, 1107.

[926] _Brut_, 1113.

[927] _Ibid._, 1135.

[928] _Ibid._, 1135, 1157, 1199, 1203, 1207.

[929] Meyrick's _Hist. of Cardigan_, p. 293. Dinerth is not the same
as Llanrhystyd, though Lewis (_Top. Dict. Wales_) says it is; the two
places have separate mention in _Brut_, 1157. Mr Clark mentions the
motte. _M. M. A._, i., 115.

[930] _Brut_, 1135.

[931] Meyrick's _Hist. of Cardigan_, p. 232.

[932] _Brut_, 1157.

[933] _Beauties of England and Wales_, Cardigan, p. 502.

[934] _Brut_, under 1113.

[935] In the _Rolls_ edition of the _Brut_ this castle is called
Llanstephan, but the context makes it probable that Lampeter is meant;
the _Annales Cambriæ_ say "the castle of Stephen."

[936] _Beauties of England and Wales_, p. 492.

[937] _Brut_, 1216.

[938] _Arch. Journ._, xxviii., 293.

[939] _Brut_, 1094.

[940] _Desc. Camb._, i., 10.

[941] _Brut_, 1094.

[942] _Ibid._, p. 110. There is a farmhouse called Rhyd y Gors about
a mile lower down than Carmarthen, and on the opposite side are some
embankments; but I am assured by Mr Spurrell of Carmarthen that these
are only river-embankments. Rhyd y Gors means the ford of the bog;
there is no ford at this spot, but there was one at Carmarthen.

[943] See _Arch. Camb._, 1907, pp. 237-8.

[944] See Round's _Ancient Charters_, p. 9, _Pipe Roll_ Series, vol. x.

[945] _Brut_, 1113.

[946] The first mention of the castle of Llanstephan is in the _Brut_,
1147, if, as has been assumed above, the mention in 1136 refers to
Stephen's castle at Lampeter, as the _Annales Cambriæ_ say.

[947] The motte of Conisburgh in Yorkshire is a very similar case known
to the writer; it measures 280 × 150 feet. Such very large mottes could
rarely be artificial, but were formed by entrenching and scarping a
natural hill.

[948] _Brut_, 1256. See _Arch. Camb._, 1907, p. 214, for Col. Morgan's
remarks on this castle.

[949] The name _Gueith tineuur_ is found in the _Book of Llandaff_, p.
78 (Life of St Dubricius), but it seems doubtful whether this should
be taken to prove the existence of some "work" at Dinevor in the 6th
century. See Wade-Evans, _Welsh Mediæval Law_, p. 337-8.

[950] _Brut_, 1145. "Cadell ap Griffith took the castle of Dinweiler,
which had been erected by Earl Gilbert."

[951] _Gwentian Chronicle._

[952] The statement of Donovan (_Excursions Through South Wales_), that
the castle stands on an artificial mount is quite incorrect.

[953] The _Rolls_ edition of the _Brut_ gives the corrupt reading Aber
Cavwy for the castle of "Robert the Crook-handed," but a variant MS.
gives Aber Korram, and it is clear from the _Gwentian Chronicle_ and
Powell (p. 145) that Abercorran is meant.

[954] _Brut_, 1152.

[955] See paper by Mr D. C. Evans, _Arch. Camb._, 1907, p. 224.

[956] The first mention known to the writer is in 1285.

[957] _Arch. Camb._, 3rd ser., v., 346.

[958] _Annales Cambriæ_, 1205; _Brut_, 1207, 1208. The _Annales_ call
it the castle of Luchewein.

[959] _Beauties of England and Wales_, "Caermarthen," pp. 192, 309.

[960] _Mon. Ang._, iii., 244.

[961] This motte is mentioned in a charter of Roger, Earl of Hereford,
Bernard's grandson, in which he confirms to the monks of St John
"molendinum meum situm super Hodeni sub pede mote castelli." _Arch.
Camb._, 1883, p. 144.

[962] The dates in the _Brut_ are now one year too early. Under 1209
it says, "Gelart seneschal of Gloucester fortified (cadarnhaaod) the
castle of Builth." We can never be certain whether the word which
is translated _fortified_, whether from the Welsh or from the Latin
_firmare_, means built originally or rebuilt.

[963] _Beauties of England and Wales_, "Brecknockshire," p. 153.

[964] _Brut_, in _anno_. The Mortimers were the heirs of the De Braoses
and the Neufmarchés.

[965] _Annales Cambriæ_, 1260. This may, however, be merely a figure of
speech.

[966] Order to cause Roger Mortimer, so soon as the castle of Built
shall be closed with a wall, whereby it will be necessary to remove the
bretasches, to have the best bretasche of the king's gift. _Cal. of
Close Rolls_, Ed. I., i., 527.

[967] See Clark, _M. M. A._, i., 307.

[968] Round, _Ancient Charters_, No. 6.

[969] _Itin._, v., 74.

[970] _Arch. Camb._, N. S., v., 23-28.

[971] "Wales and the Coming of the Normans," by Professor Lloyd, in
_Cymmrodorion Transactions_, 1899.

[972] Marchegay, _Chartes du Prieurie de Monmouth_, cited by Professor
Lloyd, as above.

[973] _Brut_, 1143.

[974] Not to be confounded with the castle of Clun in Shropshire.

[975] _Annales Cambriæ_ and _Annales de Margam._ See plan in _Arch.
Camb._, 4th ser., vi., 251.

[976] _Annales Cambriæ._

[977] Really Ty-yn-yr Bwlch, the house in the pass. Not to be
confounded with Tenby in Pembrokeshire.

[978] _Cal. of Close Rolls_, Ed. II., iii., 415, 643.

[979] See "Cardiff Castle: its Roman Origin," by John Ward,
_Archæologia_, lvii., 335.

[980] See "Cardiff Castle: its Roman Origin," by John Ward,
_Archæologia_, lvii., 335.

[981] Mr Clark thought the shell wall on the motte was Norman, and
the tower Perp. But the wall of the shell has some undoubtedly Perp.
windows. The _Gwentian Chronicle_ says that Robert of Gloucester
surrounded the _town_ of Cardiff with a wall, anno 1111.

[982] See Gray's _Buried City of Kenfig_, where there are interesting
photographs. The remains appear to be those of a shell.

[983] _Annales de Margam_, 1232.

[984] Gray's _Buried City of Kenfig_, pp. 59, 150.

[985] This information is confirmed by Mr Tennant, town clerk of
Aberavon.

[986] See Francis' _Neath and its Abbey_, where the charter of De
Granville is given. It is only preserved in an Inspeximus of 1468.

[987] _M. M. A._, i., 112.

[988] Ruperra is not quite one mile from the river Rhymney. There is
another site which may possibly be that of Castle Remni: Castleton,
which is nearly 2 miles from the river, but is on the main road from
Cardiff to Newport. "It was formerly a place of strength and was
probably built or occupied by the Normans for the purpose of retaining
their conquest of Wentlwg. The only remains are a barrow in the garden
of Mr Philipps, which is supposed to have been the site of the citadel,
and a stone barn, once a chapel." Coxe's _Monmouthshire_, i., 63.

[989] It is right to say that Colonel Morgan in his admirable _Survey
of East Gower_ (a model of what an antiquarian survey ought to be) does
not connect this mound with the old castle which is mentioned, as well
as the new castle, in Cromwell's Survey of Gower. But even the old
castle seems to have been Edwardian (see the plan, p. 85), so it is
quite possible there were three successive castles in Swansea.

[990] _Brut_, 1113.

[991] Morgan's _Survey of East Gower_, p. 24.

[992] Colonel Morgan's _Survey of East Gower_.

[993] Lewis's _Topographical Dictionary_.

[994] The passage of the river Lune in Lancashire is similarly defended
by the mottes of Melling and Arkholme.

[995] The dates given are those of the _Brut_, and probably two years
too early.

[996] Meyrick's _History of Cardigan_, p. 146.

[997] Meyrick's _History of Cardigan_, p. 146.

[998] Lewis's _Topographical Dictionary_.

[999] We do not include the castles which the Welsh _re_built. Thus
in 1194 we are told that Rhys built the castle of Kidwelly, which he
certainly only rebuilt.

[1000] Malcolm Canmore himself had passed nearly fourteen years in
England. Fordun, iv., 45.

[1001] Burton remarks: "To the Lowland Scot, as well as to the Saxon,
the Norman was what a clever man, highly educated and trained in the
great world of politics, is to the same man who has spent his days in a
village." _History of Scotland_, i., 353.

[1002] Dr Round has brought to light the significant fact that King
David took his chancellor straight from the English chancery, where he
had been a clerk. This first chancellor of Scotland was the founder of
the great Comyn family. _The Ancestor_, 10, 108.

[1003] Fordun, _Annalia_, vol. iv.

[1004] It is tempting to connect the extraordinary preponderance of
mottes, as shown by Dr Christison's map, in the shires which made up
ancient Galloway (Wigton, Kirkcudbright, and Dumfries) with the savage
resistance offered by Galloway, which may have made it necessary for
all the Norman under-tenants to fortify themselves, each in his own
motte-castle. It is wiser, however, to delay such speculations until
we have the more exact information as to the number of mottes in
Scotland, which it is hoped will be furnished when the Royal Commission
on Historical Monuments has finished its work. But this work will not
be complete unless special attention is paid to the earthworks which
now form part of stone castles, and which are too often overlooked,
even by antiquaries. The _New Statistical Account_ certainly raises
the suspicion that there are many more mottes north of the Forth than
are recognised in the map alluded to. In one district we are told that
"almost every farm had its _knap_." "Forfarshire," p. 326.

[1005] Cited by Fordun, v., 43.

[1006] Benedict of Peterborough, i., 68, R. S.

[1007] Fordun, v., 26. Bower in one of his interpolations to Fordun's
Annals, tells how a Highlander named Gillescop burnt certain wooden
castles (_quasdam munitiones ligneas_) in Moray. Skene's Fordun, ii.,
435.

[1008] That Fordun should speak of the _castra_ and _municipia_ of
Macduff is not surprising, seeing that he wrote in the 14th century,
when a noble without a castle was a thing unthinkable.

[1009] Burton actually thought that the Normans built no castles in
Scotland in the 12th century. Messrs MacGibbon and Ross remark that
there is not one example of civil or military architecture of the 12th
century, while there are so many fine specimens of ecclesiastical.
_Castellated Architecture of Scotland_, i., 63. It is just to add
that when speaking of the castles of William the Lion, they say: "It
is highly probable that these and other castles of the 13th century
were of the primeval kind, consisting of palisaded earthen mounds and
ditches." _Ibid._, iii. 6.

[1010] _Mote_ is the word used in Scotland, as in the north of England,
Pembrokeshire, and Ireland, for the Norman _motte_. As the word is
still a living word in Scotland, its original sense has been partly
lost, and it seems to be now applied to some defensive works which are
not mottes at all. But the true motes of Scotland entirely resemble the
mottes of France and England.

[1011] _Scottish Review_, xxxii., 232.

[1012] _Scottish Review_, xxxii., 232.

[1013] _Ibid._, p. 236.

[1014] This list is mainly compiled from Chalmers' _Caledonia_,
vol. i., book iv., ch. i. The letter C. refers to Dr Christison's
_Early Fortifications in Scotland_; N., to Mr Neilson's paper in the
_Scottish Review_, 1898; O.M., to the 25-inch Ordnance Map; G., to the
_Gazetteer of Scotland_. It is a matter of great regret to the writer
that she has been unable to do any personal visitation of the Scottish
castles, except in the cases of Roxburgh and Jedburgh. It is therefore
impossible to be absolutely certain that all the hillocks mentioned in
this list are true mottes, or whether all of them still exist.

[1015] _Registrum Magni Sigilli_, quoted by Christison, p. 19.

[1016] A plan is given by Mr Coles in "the Motes, Forts, and Doons of
Kirkcudbright." Soc. _Ant. Scot._, 1891-1892.

[1017] M'Ferlie, _Lands and Their Owners in Galloway_, ii., 47.

[1018] This description, taken from the _Gazetteer_, seems clear, but
Mr Neilson tells me the site is more probably Woody Castle, which is
styled a manor in the 15th century. The N. S. A. says: "There is the
site of an ancient castle close to the town, on a mound of considerable
height, called the Castle Hill, which is surrounded by a deep moat."
"Dumfries," p. 383.

[1019] _Annals_, ii., 196, cited in Douglas's _History of the Border
Counties_, 173.

[1020] Round, in _The Ancestor_, 10, 108.

[1021] Dr Christison distinctly marks one on his map, but Mr Coles says
there is no trace of one, though the name Marl Mount is preserved.
_Soc. Ant. Scot._, 1892, p. 108.

[1022] See the Aberdeen volume, p. 1092.

[1023] See Grose's picture, which is confirmed by Dr Ross.

[1024] The name Tom-a-mhoid is derived by some writers from the Gaelic
_Tom_, a tumulus (Welsh Tomen) and _moid_, a meeting. Is there such
a word for a meeting in Gaelic? If there is, it must be derived from
Anglo-Saxon _mot_ or _gemot_. But there is no need to go to Gaelic
for this word, as it is clear from the _Registrum Magni Sigilli_ that
_moit_ was a common version of _mote_, and meant a castle hill, the
_mota_ or _mons castri_, as it is often called.

[1025] Chalmers, _Caledonia_, iii., 864. Sir Archibald Lawrie, however,
regards it as doubtful whether Arkel was the ancestor of the earls of
Lennox. _Early Scottish Charters_, p. 327.

[1026] M'Ferlie, _Lands and Their Owners in Galloway_, ii., 140-141.

[1027] See plan in MacGibbon and Ross, _Castellated Architecture_, iv.,
341.

[1028] The name Maccus is undoubtedly the same as Magnus, a Latin
adjective much affected as a proper name by the Norwegians of the 11th
and 12th centuries.

[1029] Lawrie, _Early Scottish Charters_, p. 273.

[1030] MacGibbon and Ross, i., 279.

[1031] _Proceedings of Soc. Ant. Scotland_, xxxi., and N. S. A.

[1032] See Armstrong's _History of Liddesdale_, cited by MacGibbon and
Ross, i., 523.

[1033] Round, _The Ancestor_, No. 11, 130.

[1034] _Benedict of Peterborough_, i., 67. See Mr Neilson's papers in
the _Dumfries Standard_, June 28, 1899. Mr Neilson remarks: "It may
well be that the original castle of Dumfries was one of Malcolm IV.'s
forts, and that the mote of Troqueer, at the other side of a ford of
the river, was the first little strength of the series by which the
Norman grip of the province was sought to be maintained."

[1035] "Mottes, Forts, and Doons of Kirkcudbright," _Soc. Ant. Scot._,
xxv., 1890.

[1036] The _Annals of the Four Masters_ mention the building of three
castles (caisteol) in Connaught in 1125, and the _Annals of Ulster_
say that Tirlagh O'Connor built a castle (caislen) at Athlone in 1129.
What the nature of these castles was it is now impossible to say, but
there are no mottes at the three places mentioned in Connaught (Dunlo,
Galway, and Coloony). The _caislen_ at Athlone was not recognised by
the Normans as a castle of their sort, as John built his castle on a
new site, on land obtained from the church. _Sweetman's Cal._, p. 80.

[1037] The meagre entries in the various _Irish Annals_ may often
come from contemporary sources, but as none of their MSS. are older
than the 14th century, they do not stand on the same level as the two
authorities above mentioned.

[1038] "Hibernicus enim populus castella non curat; silvis namque pro
castris, paludibus utitur pro fossatis." _Top. Hib._, 182, R. S., vol.
v. In the same passage he speaks of the "fossa infinita, alta nimis,
rotunda quoque, et pleraque triplicia; castella etiam murata, et adhuc
integra, vacua tarnen et deserta," which he ascribes to the Northmen.
This passage has been gravely adduced as an argument in favour of the
prehistoric existence of mottes! as though a round _ditch_ necessarily
implied a round _hill_ within it! Giraldus was probably alluding to the
round embankments or _raths_, of which such immense numbers are still
to be found in Ireland. By the "walled castles" he probably meant the
stone enclosures or _cashels_ which are also so numerous in Ireland.
In the time of Giraldus the word _castellum_, though it had become the
proper word for a private castle, had not quite lost its original sense
of a fortified enclosure of any kind, as we know from the phrases "the
castle and tower" or "the castle and motte" not infrequent in documents
of the 12th century (see Round's _Geoffrey de Mandeville_, Appendix O,
p. 328). We may add that Giraldus' attribution of these prehistoric
remains to Thorgils, the Norwegian, only shows that their origin was
unknown in his day.

[1039] See _Expug. Hib._, 383, 397, 398.

[1040] I am informed that the "Crith Gablach," which gives a minute
description of one of these halls, is a very late document, and by no
means to be trusted.

[1041] _Vide_ the _Irish Annals_, passim.

[1042] There is another story, preserved in _Hanmer's Chronicle_, that
the Irish chief Mac Mahon levelled two castles given to him by John de
Courcy, saying he had promised to hold not stones but land.

[1043] Joyce's _Irish Names of Places_, p. 290.

[1044] See J. E. Lloyd, _Cymmrodor_, xi., 24; Skeat's _English
Dictionary_, "town." In the "Dindsenchas of Erin," edited by O'Beirne
Crowe, _Journ. R. S. A. I._, 1872-1873, phrases occur, such as "the
_dun_ was open," "she went back into the dun," which show clearly that
the _dun_ was an enclosure. In several passages _dun_ and _cathair_ are
interchanged.

[1045] Joyce, _Irish Names of Places_, p. 273.

[1046] _Annals of the Four Masters_, 1166.

[1047] See Orpen, "Motes and Norman Castles in Ireland," in _Journ. R.
S. A. I._, xxxvii., 143-147.

[1048] Sweetman's _Calendar of Documents_ relating to Ireland, i., 412.

[1049] That a motte-castle of earth and wood seemed to Giraldus quite
an adequate castle is proved by the fact that numbers of the castles
which he mentions have never had any stone defences. It may be a mere
coincidence, but it is worth noting, that there are no mottes now
at any of the places which Giraldus mentions as _exilia municipia_,
Pembroke, Dundunnolf, Down City, and Carrick.

[1050] This word must not be understood to mean that this new type
of castle was Edward's invention, nor even that he was the first
to introduce it into Europe from Palestine; it was used by the
Hohenstauffen emperors as early as 1224. See Köhler, _Entwickelung des
Kriegswesen_, iii., 475.

[1051] Newcastle, Worcester, Gloucester, and Bristol are instances.

[1052] Rhuddlan is an instance of this.

[1053] _Book of Rights_, p. 203.

[1054] It must be admitted that in the most recent and most learned
edition of the _Anglo-Saxon Chronicle_ the topographical identifications
are quite on a level with O'Donovan's.

[1055] The _Annals_ have not been used, partly because in their present
form they are not contemporary, and partly because the difficulties of
identifying many of the castles they mention appeared insuperable.

[1056] See especially two papers on "Motes and Norman Castles in
Ireland," in _English Historical Review_, vol. xxii., pp. 228, 240. Mr
Orpen has further enriched this subject by a number of papers in the
_Journ. R. S. A. I._, to which reference will be made subsequently.

[1057] The only castles still unidentified are Aq'i, Kilmehal, Rokerel,
and Inchleder.

[1058] It should be stated that the great majority of the castles in
this list have been visited for the writer by Mr Basil T. Stallybrass,
who has a large acquaintance with English earthworks, as well as a
competent knowledge of the history of architecture. The rest have
been visited by the writer herself, except in a few cases where the
information given in Lewis's _Topographical Dictionary_ or other
sources was sufficient. The castles personally visited are initialled.

[1059] _Annals of Loch Cè_.

[1060] Orpen, _Eng. Hist. Rev._, xxii., 249.

[1061] Orpen, _Eng. Hist. Rev._, xxii., 450, citing from MS. _Annals of
Innisfallen_.

[1062] The poetical list enumerates the places which were "of the
right of Cashel in its power." The prose version, which may be assumed
to be later, is entitled "Do phortaibh righ Caisil," which O'Donovan
translates "of the seats of the king of Cashel." But can one small king
have had sixty-one different abodes? Professor Bury says "The _Book of
Rights_ still awaits a critical investigation." _Life of St Patrick_,
p. 69.

[1063] _Ibid._, p. 449. See Westropp, Trans. R. I. A., xxvi. (c), p.
146. Mr Orpen informs me that the _Black Book of Limerick_ contains a
charter of William de Burgo which mentions "Ecclesia de Escluana alias
Kilkyde." No. cxxxv.

[1064] _Journ. R. S. A. I._, 1898, 155; and 1904, 354.

[1065] _Eng. Hist. Rev._, xxii., 452.

[1066] Butler's _Notices of the Castle of Trim_, p. 13.

[1067] _Eng. Hist. Rev._, xxii., 458.

[1068] _Eng. Hist. Rev._, xxii., 441.

[1069] "Exile municipium," _Giraldus_, 345. See _Eng. Hist. Rev._, xx.,
717.

[1070] _Annals of Ulster_, 1177.

[1071] See Orpen, "Motes and Castles in County Louth," _Journ. R. S. A.
I._, xxxviii., 249. The town walls are later than the castle, and were
built up to it.

[1072] Cited by Westropp, _Journ. R. S. A. I._, 1904, paper on "Irish
Motes and Early Norman Castles."

[1073] _Annals of Ulster_, 1186.

[1074] Round, _Cal. of Doc._ preserved in France, i., 105, 107.

[1075] "On the Ancient Forts of Ireland," _Trans. R. I. A._, 1902.

[1076] Orpen, "The Castle of Raymond le Gros at Fodredunolan," _Journ.
R. S. A. I._, 1906.

[1077] _Annals of Innisfallen._

[1078] Orpen, _Eng. Hist. Rev._, xxii., 449.

[1079] "On some Caves in the Slieve na Cailliagh District," by E. C.
Rotheram, _Proc. R. I. A._, 3rd ser., vol. iii. Mr Rotheram remarks
that the passages in the motte of Killallon, and that of Moat near
Oldcastle, seem as if they were not built by the same people as those
who constructed the passages at Slieve na Cailliagh.

[1080] _Annals of Ulster_.

[1081] _Annals of Loch Cè_.

[1082] Orpen, _Eng. Hist. Rev._, xxii., 448.

[1083] _Ibid._, p. 242.

[1084] _Annals of Ulster_. See Orpen, _Eng. Hist. Rev._, xxii., 443.

[1085] _Annals of Ulster_.

[1086] _Annals of the Four Masters_, vol. iii. See Orpen, _Journ. R. S.
A. I._, vol. xxxix., 1909.

[1087] Orpen, _Eng. Hist. Rev._, xxii., 448. A place called Graffan
is mentioned in the _Book of Rights_, and on the strength of this
mere mention it has been argued that the motte is a prehistoric work.
_Trans. R. I. A._, vol. xxxi., 1902.

[1088] Mr Orpen.

[1089] Giraldus' words are: "Castrum Lechliniæ, super nobilem Beruæ
fluvium, a latere Ossiriæ, trans Odronam in loco natura munito." V.,
352. See _Eng. Hist. Rev._, xxii., 245.

[1090] See Orpen, _Eng. Hist. Rev._, xxii., 456, and _Journ. R. S. A.
I._, xxxvii., 140.

[1091] Orpen, "Motes and Norman Castles in County Louth," _Journ. R.
S. A. I._, xxxviii., 241, from which paper the notice above is largely
taken.

[1092] _Eng. Hist. Rev._, xxii., 242.

[1093] The castle is casually mentioned by Giraldus, v., 100, and the
date of its erection is not given.

[1094] As far as the writer's experience goes, terraces are only found
on mottes which have at some time been incorporated in private gardens
or grounds.

[1095] _Journ. R. S. A. I._, vol. xxxix., 1909.

[1096] Piers, _Collect. de Rebus Hib._, cited by Orpen.

[1097] Mr Orpen says: "The castle was 'constructed anew' in the sixth
and seventh years of Edward I., when £700 was expended." _Irish Pipe
Rolls_, 8 Edward I., cited in _Eng. Hist. Rev._, xxii., 454.

[1098] Line 3178.

[1099] The annular bailey, with the motte in the centre, is a most
unusual arrangement, and certainly suggests the idea that the motte was
placed in an existing Irish rath.

[1100] See Appendix M.

[1101] _Annals of Loch Cè_.

[1102] _Giraldus_, v., 313.

[1103] This keep has a square turret on each of its faces instead of at
the angles. A similar plan is found at Warkworth, and Castle Rushen,
Isle of Man.

[1104] Orpen, _Eng. Hist. Rev._, xxii., 248.

[1105] Figured in _The Tomb of Ollamh Fodhla_, by E. A. Conwell, 1873.

[1106] _Gir._, i., 255, 277.

[1107] _Eng. Hist. Rev._, xxii., 457.

[1108] In five cases the mottes are now destroyed.

[1109] The dates of the building of numbers of these castles are given
in the _Annals of Ulster_ and the _Annals of Loch Cè_.

[1110] _Cal. of Pat. Rolls_, 1232-1247.

[1111] The tower at Malling was supposed to be an early Norman keep by
Mr G. T. Clark (_M. M. A._, ii., 251), but it has recently been shown
that it is purely an ecclesiastical building.

[1112] The only stone castles of early date in France which the writer
has been able to visit are those of Langeais, Plessis Grimoult,
Breteuil, and Le Mans. The two latter are too ruinous to furnish data.

[1113] Given in D'Achery's _Spicilegium_, iii., 232.

[1114] This can be positively stated of Baugé, Montrichard, Montboyau,
St Florent-le-Vieil, Chateaufort, and Chérament. M. de Salies thinks
the motte of Bazonneau, about 500 metres from the ruins of the castle
of Montbazon, is the original castle of Fulk Nerra. _Histoire de Fulk
Nerra_, 57. About the other castles the writer has not been able to
obtain any information.

[1115] See Halphen, _Comté d'Anjou au xiième Siècle_, 153.

[1116] The building of Langeais was begun in 994. _Chron. St Florent_,
and _Richerius_, 274.

[1117] It somewhat shakes one's confidence in De Caumont's accuracy
that in the sketch which he gives of this keep (_Abécédaire_, ii., 409)
he altogether omits this doorway.

[1118] Measurements were impossible without a ladder.

[1119] It is well known that William the Conqueror left large treasures
at his death.

[1120] The keep of Colchester is immensely larger than any keep in
existence. Mr Round thinks it was probably built to defend the eastern
counties against Danish invasions. _Hist. of Colchester Castle_, p.
32. Its immense size seems to show that it was intended for a large
garrison.

[1121] _Cours d'Antiquités Monumentales_, v., 152, and _Abécédaire_,
ii., 413-431. De Caumont says of the keep of Colchester, "il me parait
d'une antiquité moins certaine que celui de Guildford, et on pourrait
le croire du douzième siècle" (p. 205), a remark which considerably
shakes one's confidence in his architectural judgment.

[1122] As only the foundations of Pevensey are left, it gives little
help in determining the character of early keeps. It had no basement
entrance, and the forebuilding is evidently later than the keep.

[1123] The Tower had once a forebuilding, which is clearly shown in
Hollar's etching of 1646, and other ancient drawings. Mr Harold Sands,
who has made a special study of the Tower, believes it to have been a
late 12th-century addition.

[1124] Tiles are not used in the Tower, but some of the older arches of
the arcade on the top floor have voussoirs of rag, evidently continuing
the tradition of tiles. Most of the arches at Colchester are headed
with tiles.

[1125] The room supposed to be the chapel in Bamborough keep has a
round apse, but with no external projection, being formed in the
thickness of the wall. The keep of Pevensey has three extraordinary
apse-like projections of solid masonry attached to its foundations. See
Mr Harold Sands' _Report of Excavations at Pevensey_.

[1126] "In the course of the 12th century, the base of the walls was
thickened into a plinth, in order better to resist the battering ram."
(_Manuel d'Archæologie Française_, ii., 463.) The keep of Pevensey has
a battering plinth which is clearly original, and which throws doubt
either on this theory of the plinth, or on the age of the building.

[1127] It is well known that blocks of huge size are employed in
Anglo-Saxon architecture, but generally only as quoins or first
courses. See Baldwin Brown, _The Arts in Early England_, ii., 326.

[1128] Baldwin Brown, "Statistics of Saxon Churches," _Builder_, Sept.
1900.

[1129] Mr Round gives ground for thinking that this keep was built
between 1080 and 1085. _Colchester Castle_, p. 32.

[1130] Piper's _Burgenkunde_, p. 85.

[1131] Schulz, _Das Hofische Leben zur Zeit der Minnesinger_, i., 59.
Grose writes of Bamborough Castle: "The only fireplace in it was a
grate in the middle of a large room, where some stones in the middle
of the floor are burned red." He gives no authority. _Antiquities of
England and Wales_, iv., 57.

[1132] "The type of castle created in the 10th century persisted till
the Renascence." Enlart, _Manuel d'Archæologie_, ii., 516.

[1133] See Appendix N.

[1134] Enlart, _Manuel d'Archæologie_, ii., 516. "Jusqu'au milieu du
xii^ième siècle, et dans les exemples les plus simples des époques
qui suivent, le donjon est bien près de constituer à lui seul tout le
château."

[1135] _Abriss der Burgenkunde_, 50-60.

[1136] _Entwickelung des Kriegswesen_, iii., 352 and 428. No
continental writers are entirely to be trusted about English castles;
they generally get their information from Clark, and it is generally
wrong.

[1137] This of course explains why the castle of London is always
called _The Tower_; it was originally the only tower in the fortress.

[1138] The _Close Rolls_ mention _palicia_ or stockades at the castles
of Norwich, York, Devizes, Oxford, Sarum, Fotheringay, Hereford,
Mountsorel, and Dover.

[1139] _Close Rolls_, i., 195a and 389.

[1140] See Chapter VI., p. 89, and Appendix O.

[1141] Piper states that the evidence of remains proves that the lower
storey was a prison. But these remains probably belong to a later date,
when the donjon had been abandoned as a residence, and was becoming the
_dungeon_ to which prisoners were committed. The top storey of the keep
was often used in early times as a prison for important offenders, such
as Conan of Rouen, William, the brother of Duke Richard II., and Ranulf
Flambard.

[1142] See Appendix P.

[1143] At Conisburgh and Orford castles there are ovens on the roofs,
showing that the cooking was carried on there; these are keeps of Henry
II.'s time.

[1144] De Caumont says these remains are on a motte, a strange
statement, as they are only a foot or two above the surrounding level.

[1145] No stone castles in England are known to have been built by
William Rufus; he built Carlisle Castle, but probably only in wood. As
we have seen, several Welsh castles were built in his time, but all in
earth and timber.

[1146] Built by Archbishop William of Corbeuil. _Gervase of Canterbury_,
R. S., ii., 382.

[1147] Robert de Torigny, also called Robert de Monte, was Abbot of
Mont St Michael during the lifetime of Henry II., and was a favoured
courtier whose means of obtaining information were specially good.
French writers are in the habit of discounting his statements, because
they do not recognise the almost universal precedence of a wooden
castle to the stone building, which when it is recognised, completely
alters the perspective of castle dates. See Appendix Q.

[1148] The keep of Caen, which was square, was demolished in 1793. De
Caumont, _Cours d'Antiquités_, v., 231. The keep of Alençon is also
destroyed. There are fragments of castles at Argentan, Exmes, and
St Jean-le-Thomas. The keep of Vernon or Vernonnet is embedded in a
factory. _Guide Joanne_, p. 6.

[1149] The writer has also visited Vire and Le Mans, but even if the
walls of the keep of Vire, of which only two sides remain, were the
work of Henry I., the details, such as the corbelled lintel, the window
benches, and the loop in the basement for a crossbow, point to a later
period. At Le Mans, to the north of the cathedral, is a fragment of
an ancient tower, built of the rudest rubble, with small quoins of
ashlar; this may be the keep built by William I., which Wace says was
of stone and lime (p. 234, Andresen's edition). It is difficult to
examine, being built up with cottages. Domfront, like Langeais, is only
a fragment, consisting of two walls and some foundations.

[1150] _Dictionnaire de l'Architecture._

[1151] _M. M. A._, i., 186.

[1152] In speaking of Falaise, of course we only mean the great square
keep, and not the Little Donjon attached to it at a later period, nor
the fine round keep added by Talbot in the 15th century.

[1153] Small spaces, such as the chapel, passages, and mural chambers,
are vaulted in most keeps.

[1154] Colchester keep has only two storeys now, but Mr Round argues
that it must have had three, as a stairway leads upward from the second
floor, in the N.W. tower, and some fragments of window cases remain as
evidence. _Colchester Castle_, p. 92.

[1155] The Tower and Colchester keep both have wells, which are seldom
wanting in any keep. There was no appearance of a well at Langeais, but
excavation might possibly reveal one.

[1156] The first castle at Corfe was built by William's half-brother,
Robert, Count of Mortain. The keep of Corfe is sometimes attributed
to him, but when we compare its masonry with that of the early hall
or chapel in the middle bailey, we shall see that this date is most
unlikely. Norwich was always a royal castle.

[1157] Part of the basement of Norwich keep has pillars, from which it
has been assumed that it was vaulted; but no trace of vaulting is to be
seen.

[1158] The only decoration at Corfe keep is in the oratory, which being
at a vast height in one of the ruined walls is inaccessible to the
ordinary visitor. Corfe was so much pulled about by Sir Christopher
Hatton in Elizabeth's reign, and is now so ruinous, that many features
are obscure. Norwich has suffered greatly from restorations, and from
re-casing.

[1159] In 1184 Henry II. paid "for re-roofing the tower of Gisors."
_Rotuli Scacc. Normanniæ_, i., 72.

[1160] It should be remembered that rude work is not invariably a sign
of age; it may only show haste, or poverty of resources. It should also
be mentioned that in the _Exchequer Rolls of Normandy_ there is an
entry of £650 in 1184 for several works at Gisors, including "the wall
round the motte" (murum circa motam). Possibly this may refer to a wall
round the foot of the motte, which seems still to exist. The shell wall
of Gisors should be compared with that of Lincoln, which is probably of
the first half of the 12th century.

[1161] No decagonal tower of Henry I.'s work is known to exist; all his
tower keeps are square.

[1162] Bower, _Scotichronicon_, v., 42. This passage was first pointed
out by Mr George Neilson in _Notes and Queries_, 8th ser., viii., 321.
The keep of Carlisle has been so much pulled about as to obscure most
of its features. The present entrance to the basement is not original.

[1163] _M. M. A._, i., 353.

[1164] Unfortunately the greater part of these valuable _Rolls_ is
still unpublished. The Pipe Roll Society is issuing a volume every
year, and this year (1910) has reached the 28th Henry II.

[1165] The keeps of Richmond and Bowes were only finished by Henry II.;
Richmond was begun by Earl Conan, who died in 1170, when Henry appears
to have taken up the work. Bowes was another of Earl Conan's castles.
Tickhill is now destroyed to the foundations, but it is clear that it
was a tower. The writer has examined all the keeps mentioned in this
list. It will be noticed that most of the towers took many years to
build.

[1166] Henry built one shell keep of rubble and rag, that of Berkeley
Castle, which is not mentioned in the _Pipe Rolls_, having been built
before his accession. It is noteworthy that he did not build it for
himself, but for his ally, Robert Fitz Hardinge.

[1167] The basement storey of Chester keep (the only part which now
remains) is also vaulted, but this can scarcely be Henry's work, for
though he spent £102 on this castle in 1159, it must have been begun by
Ranulf, Earl of Chester, in Stephen's reign. Moreover, it is doubtful
whether the vaulting, which is covered by whitewash, is really ancient.

[1168] Leland says of Wark, "the dongeon is made of foure howses
hight," but probably he included the basement.

[1169] The earliest instance of a portcullis groove with which the
writer is acquainted is in the basement entrance of Colchester. It is
obvious to anyone who carefully examines this entrance and the great
stair to the left of it that they are additions of a later time than
William's work. The details seem to point to Henry I.'s reign. The keep
of Rochester has also a portcullis groove which seems to be a later
addition.

[1170] King, paper on Canterbury Castle in _Archæologia_, vi., 298. We
have not observed in any English keeps (except in this single instance)
any of the elaborate plans to entrap the enemy which M. Viollet le Duc
describes in his article on Donjons. He was an imaginative writer, and
many of his statements should not be accepted without reserve.

[1171] Wark was also an octagonal keep, but there is considerable
doubt whether this octagonal building was the work of Henry II., as
Lord Dacre wrote to Wolsey in 1519 concerning Wark that "the dongeon
is clerely finished," and mentions that all the storeys but one were
vaulted with stone. This makes it almost certain that the castle of
Wark was entirely rebuilt at this time, after having been demolished by
the Scots in 1460. It is now an utter ruin, and even the foundations of
the keep are buried.

[1172] At Thorne, near Doncaster, where the great earls Warenne had a
castle, there are the foundations, on a motte, of a keep which seems to
resemble that of Orford; it ought to be thoroughly excavated.

[1173] These measurements are from Grose, _Antiquities_, v., 74.

[1174] See Payne Gallwey, _The Crossbow_, 309; Köhler, _Kriegswesen_,
iii., 192. The trébuchet is first mentioned at the siege of Piacenza in
1199.

[1175] As far as we can tell, the tops of keeps having generally been
ruined or altered, the common arrangement was either a simple gable,
or two gables resting on a cross wall, such as all the larger keeps
possessed.

[1176] Another consequence of the introduction of an engine of longer
range was the widening of castle ditches. We frequently find works on
ditches mentioned in John's accounts.

[1177] Payne Gallwey, _The Crossbow_, p. 3. We find it used by
Louis VI. of France, before 1137. Suger's _Gesta Ludovici_, 10 (ed.
Molinier). Ten balistarii are mentioned in Domesday Book, but they may
have been engineers of the great balista, a siege machine. There is no
representation of a crossbow in the Bayeux Tapestry. There are entries
in the _Pipe Rolls_ of 6, 8, and 9 Henry II. of payments for arbelast,
but these also may refer to the great balista.

[1178] _Guill. Brit. Armorici Philippides_, Bouquet xvii., line 315.

[1179] The bow brought by Richard from Palestine is believed to have
been an improved form of crossbow, made of horn and yew, "light,
elastic, and far more powerful than a bow of solid wood." Payne
Gallwey, _The Crossbow_.

[1180] "Fenestris arcubalistaribus," Bouquet xvii., 75. The writer has
never found a single defensive loophole in any of the keeps of Henry I.
or Henry II. Köhler remarks that the loopholes up to this period do not
seem to be intended for shooting (_Entwickelung des Kriegswesen_, iii.,
409), and Clark has some similar observations.

[1181] _Dictionnaire de l'Architecture_, art. "Meurtrière."

[1182] Meyrick in his _Ancient Armour_ quotes a charter of 1239,
in which the French king grants a castle to the Count de Montfort
on condition "quod non possumus habere in eodem archeriam nec
arbalisteriam," which Meyrick audaciously translates "any perpendicular
loophole for archers, nor any cruciform loophole for crossbowmen." The
quotation is unfortunately given by Sir R. Payne Gallwey without the
Latin original. It is at any rate probable that the cruciform loophole
was for _archers_; it does not appear till the time of the long-bow,
which was improved and developed by Edward I., who made it the most
formidable weapon of English warfare.

[1183] See Appendix H.

[1184] _Entwickelung des Kriegswesen_, iii., 417.

[1185] In 1186, the Duke of Burgundy caused the towers and walls of
his castle of Chatillon to be "hoarded" (hordiari). This duke had been
a companion of Richard's on the third crusade. William le Breton,
_Philippides_, line 600. Richard's _hurdicia_ at Chateau Gaillard were
two years earlier.

[1186] See Dieulafoy, _Le Chateau Gaillard et l'Architecture Militaire
au Treizième Siècle_, p. 13.

[1187] The best French and German authorities are agreed about this.
The holes in which the wooden beams supporting the hurdicia were placed
may still be seen in many English castles, and so may the remains of
the stone brackets. They would be good indications of date, were it not
that hurdicia could so easily be added to a much older building.

[1188] Köhler gives the reign of Frederic Barbarossa (1155-1191) as the
time of the first appearance of the round keep in Germany.

[1189] In spite of this, I cannot feel satisfied that the keep of
Étampes is of so early a date. The decorative features appear early,
but the second and third storeys are both vaulted, which is a late
sign. The keep called Clifford's Tower at York, built by Henry III.
1245 to 1259, is on the same plan as Étampes.

[1190] This keep has been long destroyed.

[1191] Ground entrances occur in several much earlier keeps, as
at Colchester (almost certainly an addition of Henry I.'s time),
Bamborough (probably Henry II.'s reign), and Richmond, where Earl Conan
seems to have used a former entrance gateway to make the basement
entrance of his keep. See Milward, _Arch. Journ._, vol. v.

[1192] Built by Earl Hamelin, half-brother of Henry II., who died in
1201.

[1193] Viollet le Duc, art. "Donjon."

[1194] The walls of the Tower are from 12 to 15 feet thick at the base;
those of Norwich 13; the four walls of Dover respectively, 17, 18,
19, and 21 feet; Carlisle, 15 feet on two sides. (Clark.) William of
Worcester tells us that Bristol keep was 25 feet thick at the base!
_Itin._, p. 260.

[1195] See Enlart, _Manuel d'Archæologie Française_, ii., 526.

[1196] MacGibbon and Ross, _Castellated Architecture of Scotland_, p.
159.

[1197] This type of castle was probably borrowed from the fortifications
of Greek cities, which the Crusaders had observed in the East.

[1198] Conway and Carnarvon consist of two adjoining courts, without
any external enclosure but a moat. Flint has a great tower outside
the quadrangle, which is sometimes mistakenly called a keep, but its
internal arrangements show that it was not so, and it is doubtful
whether it was ever roofed over. It was simply a tower to protect the
entrance, taking the place of the 13th-century barbican.

[1199] Köhler states that the gatehouse palace is peculiar to England:
"only at Perpignan is there anything like it." _Entwickelung des
Kriegswesen_, iii., 480.

[1200] Köhler mentions the castle of Neu Leiningen as the first example
in Germany, built in 1224. _Kriegswesen_, iii., 475. Frederic II.'s
castles were of this type. The castle of Boulogne, finished in 1231,
is one of the oldest examples of the keepless type in France. Enlart,
_Archæologie Française_, ii., 534. The Bastille of Paris was a castle
of this kind. According to Hartshorne, Barnwell Castle, in Northants,
is of the keepless kind, and as the _Hundred Rolls_ state that it was
built in 1264, we seem to have here a positive instance of a keepless
castle in Henry III.'s reign. _Arch. Inst. Newcastle_, vol. 1852. And
it appears to be certain that Gilbert de Clare, Earl of Gloucester,
built the keepless castle of Caerphilly before Edward came to the
throne. See Little's _Mediæval Wales_, p. 87.

[1201] French archæologists are enthusiastic over the keep of Chateau
Gaillard, the scientific construction of the towers of the curtain, the
avoidance of "dead angles," the continuous flanking, etc. See Viollet
le Duc, art. "Chateau," and Dieulafoy, _Le Chateau Gaillard_.

[1202] This type is extremely rare: Trim, in Ireland, and Castle
Rushen, in the Isle of Man, are the only other instances known to the
writer. Trim is a square tower with square turrets in the middle of
each face; Castle Rushen is on the same plan, but the central part
appears to have been an open court.

[1203] Enlart, _Archæologie Française_, ii., 516.

[1204] Martène's _Thesaurus Anecdotorum_, iv., 118. "Nulli licuit in
Normannia fossatum facere in planam terram, nisi tale quod de fundo
potuisset terram jactare superius sine scabello. Et ibi nulli licuit
facere palicium, nisi in una regula; et id sine propugnaculis et
alatoriis. Et in rupe et in insula nulli licuit facere fortitudinem, et
nulli licuit in Normannia castellum facere."

[1205] The document which calls itself _Leges Henrici Primi_, x., 1,
declares the "castellatio trium scannorum" to be a right of the king.
_Scannorum_ is clearly _scamnorum_, banks. It is noteworthy that a
motte-and-bailey castle is actually a fortification with three banks:
one round the top of the motte, one round the edge of the bailey, one
on the counterscarp of the ditch.

[1206] See the case of Benhall, _Close Rolls_, ii., 52b (1225).

[1207] Aldreth and Burton are omitted from this list.

[1208] M. and B. stand for Motte and Bailey; K. and B. for Keep and
Bailey; O. for Outside the Town.




    Transcriber's Note:

    Every effort has been made to replicate this text as faithfully as
    possible. The errata listed at the beginning of the book have been
    fixed, and some minor corrections of spelling and punctuation have
    been made.

    Italic text has been marked with _underscores_.
    Bold text has been marked with =equals signs=.
    Letters preceded by a ^caret appeared as superscripts to the end of
    the word.
    OE ligatures have been expanded.