Oxford in Ingulph's description of Crowland. However, from the fact that no MS.
earlier than the close of the sixteenth century is in existence, this gratuitous inter-
polation cannot be brought home to him.
^ Orderic Vital, book iv. cap. 4 ; V>\xQ!a.t'aa&, Hist. Norm. Scr-ipt.i^. ^\o. Orderic
Vital was born at Atcham, near to Shrewsbury, in 1075. At five years of age he
went to Shrewsbury to school. At ten years old he went over to the monastery of
St. Evroult, in Normandy, where he lived the greater part of his life. He certainly
on one occasion, and probably on more than one during his sojourn there, visited
England. In 11 15 he tells us he spent some days at Crowland, in Lincolnshire;
but most likely it was on another occasion that he went to Worcester and the
neighbourhood of his birth. His father died in mo. All this shows that he
might well have conversed with those who had been present as young men at the
siege of Exeter in 1067, and hence his story may be relied upon.
198 THE EARLY HISTORY OF OXFORD.
evidence bearing on the subject of Oxford being besieged. This is
Roger of Wendover's chronicle. The passage as it appears in the
only MS. existing is as follows :—
' Honu King William besieged Exeter and took it.
'At this time King William laid siege to the city of Exeter, which
was in rebellion against him Wherefore William being roused
to anger, with very little effort subdued the City. Thence marching
to York he almost destroyed the city \'
He is evidently summarizing William of Malmesbury : this is shown
not only by his introducing the same anecdote which that writer had
done, but also by the general context.
All then, up to this point, is quite clear, and all the chronicles follow
on one after the other, naming the two places Exeter and York, and
those two only.
Roger of Wendover's chronicle formed the basis of what is known
as Matthew Paris' Chronica Majora, and there is preserved in the
Library of Corpus College, Cambridge ^ a transcript of Roger of Wen-
dover's chronicle with additions throughout the early part down to
1235, and a continuation afterwards. The additions and the con-
tinuation, there is every reason to think, are in Matthew Paris' own
handwriting. But the transcriber, in copying Roger of Wendover,
had written Oxonia instead of Exonia, all the rest being accurately
followed. The error was not detected, and it was copied off, with
Matthew Paris' corrections, into the fine MS. preserved in the
Cottonian Library and the less important in the Harleian Collection ''.
And since the more complete copy by Matthew Paris of the St. Alban's
Chronicle became the basis of successive chronicles, the correct
reading in Roger of Wendover's original copy was entirely overlooked,
and the erroneous reading, which passed under Matthew Paris' autho
rity, found its way into all the later chronicles which treat of this period *.
^ Roger of Wendover, Chronica, sive Flares Historiarum, Eng. Hist. Soc, 184:
vol. ii. p. 4. The MS. of this chronicle is preserved in the Bodleian Library (Douce,
MS. CCVII), and is a fine vellum copy, written in the thirteenth century. After
the year 1235 occur the words, Hue tisquc scripsit cronica doniimis Rogeriis de
Wendover. This does not prove it to be the original autograph, but if it is not
it is certainly a very early transcript. The other MS., which was in the Cottonian
Collection (Otho, B. V.), was burnt, and only fragments remain. Roger of Wendover
is found to have died 1236. Appendix, § 85.
^ The MS. is known as C. C. C. C. 26. The continuation of the same as
C. C. C. C. 16. Matt. Paris, Ciironica Majora, Rolls Series, ed. 1872, vol. i. p. 465.
' The Cottonian MS. is marked Nero, D.V.; the Harleian MS. is numbered 1620.
It is noted however by Sir Frederick Madden (Matt. Paris, Historia Minor, Rolls
Series, 1866, vol. i. p. 10) that in the Cottonian MS. Exonia/n is retained in the
rubric although Oxoniain has been followed in the text.
* Matthew Paris, when compiling his Historia Angloruni (which, because it is
OXFORD AFTER THE NORMAN CONQUEST. 199
Having disposed of the only vestiges of evidence of any recorded
siege, it remains to say a word or two as to an unrecorded siege.
It is of course impossible to prove that such did not occur. But
a consideration of the circumstances renders it, a priori, highly im-
probable that Oxford was besieged by William at all. The coronation
of William at Westminster, although it virtually made him king over all
England, may not certainly have rendered him actually so. There
were the outlying districts, no doubt, which were in a state of rebellion,
and Devonshire and Cornwall seem to have found leaders to refuse
submission to the new king ; while Eadvvine and Morkere, playing, as
they did, fast and loose with W^illiam, at one moment his guest in his
Normandy progress, and the next in open rebellion against him, seem
to have gathered together a force of some kind in Northumbria early
in 1068. But the Midland counties had no rulers; as already said,
Gyrth was slain, and Leofwine also, who might have done some
service in Kent and Essex. There was no one to lead a rebellion,
and for a solitary city to stand out would have been useless with the
prestige which William had gained by his energy and decision.
It has already been pointed out that no reason can be assigned for his
besieging Oxford on his way during his first campaign in Devonshire,
when Exeter was besieged, nor in that of the north, which followed
sometime after ; it may be added, that there is no reason which can
be adduced why, in his second campaign into Yorkshire, in 1070, he
should stop to besiege a city like Oxford ; nor indeed in any of the
campaigns previous to 107 1, when we find Robert D'Oilgi in quiet
possession of the city.
The erroneous reading of Oxford, however, has permeated, as has
been said, nearly all histories, and it is necessary here to refer to a
remarkable instance in which this erroneous reading is made to
support a theory, while the theory is supposed to prove the integrity
of the reading.
In the edition of the Domesday Survey, printed by order of the English
Government in 1816, the preface by Sir Henry Ellis has the following
an abridgTuent of the Clironica Majora which he had edited, is called for con-
venience Historia Minor), follows the reading of Oxonia, not having detected
the error of the scribe. Historia Anglorum, Rolls Series, 1866, p. 10. Amongst
the later editions, so to speak, of the St. Alban's Chronicle, that which was
completed at Westminster, and which, because it incorporated Matthew Paris'
Chronicle, seems to have been attributed to an imaginary Matthew of Westminster,
has been very extensively used by the historians of the fifteenth century ; and as
that had the erroneous reading, it may be said literally to have found its way into
every English history which refers to the siege of a town at this time.
300 THE EARLY HISTORY OF OXFORD.
argument on the question of a large number of houses being returned
vastae ei desiruciae : —
* The extraordinary number of houses specified as desolated at
Oxford requires explanation. If the passage is correct, Matthew
Paris probably gives us the cause of it under the year 1067, when
William the Conqueror subdued Oxford on his ivay to York '.'
It may be asked reasonably, if this is not so, how is so unusual a
number of houses wasted and destroyed to be accounted for ? In the
first place it must be taken into account that the term vastae does not
necessarily mean destroyed, but simply empty, i.e. untenanted, and
therefore not hable to pay tax ; and houses in this state may have
made up a large proportion of the total number, 478. Many, too,
from being uninhabited, would be out of repair also. The word
desiruciae, however, is also added by the compiler of the Survey, and
therefore we ought to look for some definite act of violence. We have
not to look far for this amongst recorded events. The rebel army,
headed by Eadwine and INIorkere, marching southwards and obliging
the Gemot to be transferred from Northampton to Oxford on October
28th, 1065, as already described ^ would account for a destruction
such as this. The few words of the Chronicle give an insight into
the nature of this so-called army, in reality a rebel mob. They had
slain all the household men of Earl Tostig — that is, all men in autho-
rity and probably all who had property — and had taken all his weapons
which were at York, besides all the treasure they could lay their hands
on. They had gathered as they went southward men of Nottingham-
shire, Derbyshire, and Lincolnshire, till they came to Northampton.
Here Eadwine met them with his men, and many Welshmen, we read,
came with him. It will be remembered also that the Chronicle of
this year adds that ' the Ryihretiafi' or the ' northern men,' ' did
great harm about Northampton, while Harold went on their errand,
• General Introduction to the Domesday Survey, by Sir Henry Ellis : London,
1816, folio, p. Ixii. ; 8vo, 1833, i.p. 194. The suggestion that Matthew Paris implies
that it was on the way to York [in 1068] is distinctly erroneous, as has already beea
pointed out. Exeter and York are described in the Anglo-Saxon Chronicle, and by
the many historians who follow it, including William of Malmesbury, as two places
besieged distinctly at different times, the one before Easter, the other after Whitsun-
tide. As Matthew Paris' Chronicle is really only a transcript of that of Roger of
Wendover, which follows William of Malmesbury and has Exeter, it is unreason-
able to imply that he omitted all reference to the first campaign, and inserted
an account of a siege in the second, which no chronicler had previously ever
heard of. Thierry, as already shown, evidently follows the same lines — misled
probably by Sir Henry Ellis. In fact when once an error of the kind has been
made, all historians seem to follow it.
^ See ante, p. 181.
OXFORD AFTER THE NORMAN CONQUEST. 201
inasmuch as they slew men, and burned houses and corn, and took
all the cattle '.' When to this is added, from the contemporary life
of Edward the Confessor, already noticed, that the mob came past
the middle of England as far as Oxford ^ which agrees with the cir-
cumstance mentioned in the Saxon Chronicle of the Gemot being
finally held at Oxford, there are ample means of accounting for the
devastation which took place there. The circumstance, too, of a
large number of houses being destroyed does not point so much to
the results of a siege in the case of Oxford, where the castle stood at
one extremity of the town, as it would in the case of a town where
the castle stood in the midst, and where the houses had grown up
round it. The earthwork of King Ethelred, which was, perhaps,
sufficient to withstand the irregular forces of the Danes, would not
have availed long against the well-drilled army and the well-armed
archers of the Duke of Normandy ; and when it was taken, though
the soldiers might, out of wanton mischief, have burned some few
houses, it would not have been at all in accordance with Duke
William's policy to have allowed them to destroy the town, and
therefore it is unreasonable to assume that it was done; espe-
cially, too, as this reason is not given in the Domesday Survey, which
it probably would have been, judging from other similar incidental
notes, had the siege been the cause of the destruction of houses.
But the rebel mob of the North, joined as they were by Welshmen, and
having cast off all restraint and discipline ^, would, on arriving at a
town, be readily prompted to any wanton mischief or atrocity, and be
quite capable of destroying two-thirds of the buildings ; and though it
had happened more than twenty years before the Survey, still, remem-
bering the unsettled state of the kingdom, it is no wonder that the
men of Oxford had not repaired the losses. Those who were driven
out from their homes could not well have returned while Eadwine was
still lord over the shire *, for many who had houses in Oxford were
* Anglo-Saxon Chronicles D, E, sub anno 1065. See ante, p. 183.
^ The words describing their course are as follow: 'Nam conglomerati in infini-
tum numerum more turbinis sen tempestatis hostili expeditione perveniunt ad
Axonciiorde oppidum, satis scilicet pervagati ultra mediae Anglic terminum.' Lives
of Edward the Confessor, Rolls Series, 1858, p. 422.
^ ' Ejecto autem eo, ad vomitum reversi sunt veteris malitia?, amissoqtie freno
discipline!:, furorem adoriuntur majoris insanise.' Ibid. p. 422.
* In the last chapter it was implied that there was much difficulty in assigning
the various counties to the various earldoms. In 105 1 it may be taken as certain
that both Oxfordshire and Berkshire were included within the earldom of Swegen.
(See Florence of Worcester, sub anno.) "Whether or not for any reason Gyrth or
Leofwine had their territories extended, and either of them took in Oxfordshire,
%02 THE EARLY HISTORY OF OXFORD.
those who possessed land in the immediate neighbourhood ; for the
houses were not possessed only by the citizens who had no other
homes. During the years 1067 to 107 1 everything relating to the
security of this part of England was uncertain. Then Robert D'Oilgi
was made governor of the town, and those who had left, even sup-
posing they might have returned to their lands under new lords,
might have not cared to return to their Oxford houses even if they
had the money to restore them, which is not at all probable. On the
question, however, of the waste mansions more will have to be said
in the next chapter under the account of the Domesday Survey of
Oxford.
The next great event is the new fortification of Oxford : of this we do
not find any notice in the historians on which chief reliance has been
hitherto placed. The series of Anglo-Saxon Chronicles, as has been
pointed out, are now reduced practically to the single record supposed
to have been compiled originally, at Peterborough, but whether con-
tinued there after the Conquest is not ascertainable. Other chronicles
however, preserved and continued in difi"erent abbeys, in a measure
take their place, and, while giving a general summary of events culled
from the writings of whatever historian the chronicler happened to
possess, record here and there local events, either derived from actual
knowledge or deduced from charters or entries in registers found in
the archives of the abbey. Such, as regards Oxford, are the
Chronicles of Oseney and the Chronicle of Abingdon. Unfortunately,
no chronicle seems ever to have been kept at S. Frideswide's, or the
material for the history of Oxford might have been less scarce than it
is, nor yet at Ensham, the charters of which abbey, so far as they are
preserved, throw hardly any light upon this period.
It is to the first of these that we owe the mention of the building of
the Castle. The entry is very brief, as follows : —
'mlxxi. The same year was built {aedificatum est) the Castle of
Oxford, by Robert d'Oili the First 1.'
may be doubtful. It seems certain, however, that the Northumbrian mob with
Eadwine at their head overran Oxfordshire in 1065, and annexed it practically to
the one great northern kingdom, Harold being driven below Thames, i. e. into the old
Wessex. The Northumbrian earls seem to have overrun Mercia just as the Wessex
king had overrun Mercia more than three hundred years previously {ante, p. loS).
^ Annales Monasterii de Oseneia. Printed in Aiziiales Monastici, Rolls Series,
1869, vol. iv. p. 9. The MS. is in the Cottonian Collection, and marked Tiberius
A. 9. It is written in the same handwriting down to the year 1233, and then
continued by difterent hands ; but although this is the date of the M.S., there is no
doubt but that, generally speaking, the events have been recorded at an earlier
date. In a MS. in Corpus Christi College, Oxford, the building of the castle is
put under the year 1072. ' Robertus de Oili struxit castellum Oxonii,' Dugdale,
OXFORD AFTER THE NORMAN CONQUEST. 203
The abbey in which these annals were kept was not founded till
1 1 29, but then the founder was Robert D'Oilgi, the nephew of the
great Robert D'Oilgi, mentioned in the extract. It is natural, there-
fore, that the deeds of the uncle should be recorded in the annals of
the abbey ; besides which, the documents which came into their
possession are found incidentally to have recited the building of the
Castle.
But the question which suggests itself here is the force of the word
* built.' It does not necessarily exclude the fact of a castle existing
here before, because we know that there must have been such ; nor,
on the other hand, does it necessarily imply that he erected a castle
such as is usually conceived by the word, namely, a keep ' with stone
walls and stone towers surrounding it : but there is a middle course
between the two which may reasonably be taken. We were not
indebted to him wholly for the Casde, nor did he make what he
found into such a castle, as we can picture, from the details and
descriptions w^hich have come down to us, to have existed in the
twelfth or thirteenth centuries. The great mound was certainly there
already ; this is not of the character of the work of the Normans at
this period ; but no doubt he deepened the ditches, and perhaps on the
west slightly extended the enceinte, and added, possibly, new palisades,
if not walls. But the main work, which struck so much the annalist,
and prompted him to use the word ' built,' was the great tower, and
that built 0/ stone, which is now existing, and which, situated upon the
line of enceinte, guarded the western approach to the Castle, The
means of attack had improved during the past hundred and fifty
years, and a lofty tower had great advantages over the mound as a
means of defence : it was less easily assailed, the defenders could
more safely reach the summit, and when there they had a much
better position against the assailants below than from the sloping
Mon. vol. vi. p. 251 ; and this agrees with a passage which occurs in the Oseney
Cartulary, from which, no doubt, it was derived.
^ It has been thought that Robert D'Oilgi might have erected something of the
nature of a stone keep on the top of the mound. If so, however, all traces of it,
even dovm to the foundation, would have been removed in Henry the Third's reign,
when the well-room was constructed at the top of the mound. There were a few
traces existing some years ago of what appeared to have been the foundation of a
tolerably large building, some fifty-eight feet in diameter, and in the form of
a decagon surrounding the hexagonal plan of the well-room, and probably of the
same date; they are laid down on the plan given in King's Vestiges of Oxford
Castle, London, 1796. The probabilities are that the builders of Robert D'Oilgi's
castle would not have ventured to erect one of the great solid structures common
in the eleventh and twelfth centuries on the top of an artificial mound of earth ;
they would have known that the foundations must soon have given way.
304 THE EARLY HISTORY OF OXFORD.
sides of a mound. It is impossible to conceive that the two were the
work of the same age, or part of the same system of fortification ; and
if so, there is no doubt the mound was the earher. But just as there
are grounds, which have already been given ', for believing that this
mound was of the early part of the tenth century, from its similarity
to those of Warwick, Tamworth, &c., which were part of one system
of fortification then adopted, so the masonry and such architectural
details as exist in the present tower leave little or no doubt but that
what we still see is the work of Robert D'Oilgi, referred to in the
Oseney Annals as having been completed in 1071.
The building of the Castle was necessitated now, not by fear of
foreign invasions, nor, indeed, of the attacks of one kingdom or
earldom by another, but by the danger of revolt. King William
knew full well that there was still an English spirit slumbering, and
that any day circumstances might arise or leaders be found by which
it might be awakened and cause him much trouble and expense to
suppress it. Numerous instances of the disturbed state of the country
may be found ; and William, besides requiring safe retreats for his
garrisons, required also prisons for those who were suspected of
treason ^ His plan seems to have been to erect castles, and confide
them to friends or followers whom he could trust. Referring to what
was done in this district the Abingdon Chronicler writes : —
' Then castles were built for the preservation of the kingdom, at
Wallingford (Walingajorde), and at Oxford {Oxeneforde), and at Windsor
{Wildesore), and at other places ^.'
These three were especially selected to guard the passage of the
Thames. We have no record of the exact date of the building of
Wallingford Castle *, and it must be remembered that the Abingdon
^ See ante, p. 117.
* The Abingdon Chronicler supplies one or two illustrations : e. g. he dilates
upon the unfortunate state of England, and first records the capture of Bishop
^gelwin of Durham [Bp. 1056-71], who, having been found in arms, was sent as
a prisoner to be kept at Abingdon Abbey : while on the other hand Ealdred, abbot
of Abingdon, who was suspected, was sent to be kept as a prisoner in Wallingford
Castle [1070-71] till he was handed over to the care of Walchelin, Bishop of Win-
chester [Bp. 1070-98], (vol. i. p. 4S6). Also, when Abbot Adelelm first came to
the abbey [c. 1071], he never went about imless accompanied by armed men ^vol. ii.
P- 3).
^ Chron. Monast. de Abingdon, Rolls Series, 1858, vol. ii. p. 3.
* It may be safely assumed that the castle built by the Conqueror at Wallingford
was situated near the river at the northern extremity of the town, where there exists
an artificial mound of the same character as that at Oxford, Warwick, Tam-
worth, &c. ; this mound was probably erected at about the same time, and for
the same purpose, as the others, though the erection is not recorded, and the
remains are not sufficient to show what William added. In after years the
OXFORD AFTER THE NORMAN CONQUEST. 205
chronicler is rather summarizing events than recording them, since
there is no reason to suppose his Chronicle to be much earlier than the
two IMS. transcripts we possess of it, namely, of the thirteenth century.
As regards Windsor, too, it is not at all clear when the Conqueror
commenced erecting the castle, upon the lofty outlier of chalk which,
surmounted as it is by the modernized medieval buildings, forms so
conspicuous an object in this part of the Thames valley. Henry of
Huntingdon records that the first time the king's court was held on this
hill was by Henry I. in mo, implying that Henry and not William
erected the same, and that all events previously chronicled as taking
place at Windsor were at Old Windsor in the parish of Clewer ^
It is singular, perhaps, that neither Orderic Vital, nor yet other his-
torians of the twelfth century make any mention of the erection of a
castle at Oxford, though they record the building of castles at several
other places ; so that were it not for the local information derived
medieval castle which took the place of William's work played an important part
at several periods of our history, notably in King Stephen's reign, in King John's
reign, and in that of Edward II. It was of considerable extent ; Leland, describing
it in Henry the Eighth's time, writes : ' The castelle yoinith to the North Gate of
the Toime, and hath 3 Dikis, large and deap, and welle waterid. About ech of
the 2 first Dikis as apon the crestes of the creastes of the Ground cast out of rennith
an embatelid Waulle now fore yn ruine, and for the most part defaced. Al the
goodly Building with the Tourres and Dungeon 3 be within the 3 Dike.' Leland's
Itinerary, Hearne's ed. vol. ii. p. 13. Camden, writing in Elizabeth's reign, also
describes it : ' Its size and magnificence used to strike me with astonishment, when
I came thither a lad from Oxford, it being a retreat for the students of Christ Church.
It is environed with a double wall and a double ditch, and in the middle on a high
artificial hill stands the citadel, in the ascent to which by steps I have seen a well
of immense depth.' (Camden's Britannia, Gough's ed., 1789, vol. i. p. 148.) The
castle again played a part in the history of the civil war in the time of Charles I,
and eventually, by an order in Council dated November 1 8, 1 65 2, it was demolished ;