shire ; hence it was necessary to distinguish the two Axe rivers.

^ Henry of Huntingdon, in giving the list of British cities attributed to Nennius
in referring to ' Kaer-legion,' writes, ' Nunc autem vix moenia ejus comparent ubi
Usca (v. 1. Asia) cadit in Sabrinam' (AIo)i. Hist. Brit. p. 692).

* We find the two forms shown on the continent, e. g. the "laap of Ptolemy and
Strabo and Isara of mediaeval writers (mod. Isere, France ; also the Isar, a tributary
of the Danube) ; the Oesia,or Oisa, of the Antonine Itinerary (mod. Oise. France) :
on the other the Escamus and Oescus of Pliny and "Ogkios of Thucydides (now
Ischa and Isker, tributaries of the Danube) ; and the Axona of Caisar (mod. Aisne,
a confluent of the Oise, France). By taking in names of places derived from the
rivers a long list might be made. It may be noticed as an example that the
Abbey of Essay in Normandy is written in the charters Axa, and Axiacum. The
word Ax, however, in some cases seems to be derived from the Latin Aquae, or
rather Aquis (e. g. Dax, Aix-la-Chapelle, Aix-les- bains, &c.), and not from the
hardening of the Celtic word.

5 The Antonine Itinerary may be of any date between the early part of the
third century and the close of the fourth century. The Iter No. XII, which de-
scribes the journey from Winchester into Wales, gives Isca Bwnnunionim, and
then, after some three or four stations, gives Iscae Leg. II. Augusta (v. 1. Iscaeleia
Atignsti, Iscalegi Augusti), which must be identified with Caer-leon on Usk. The
Iter No. XV. is practically a repetition of the above, ending with Isca Dum-
mcniorum. There can be little doubt this must be identified with Exeter.
Ptolemy in his geography {circa A. D. 1 20) seems to have confused the two, having
under ' the cities of the Dumnonii,' added after the name Isca words which belonged
to the other, as if the two cities were in one part of Britain. His list runs, "loKa, —
A a



354 THE EARLY HISTORY OF OXFORD.

at the time when their arms reached these parts the names of the two
rivers had, as spoken by the inhabitants, the same sound. But the Isc
could only have been one of the dialectic forms amongst the Britons,
because we find both in the east of Britain and the north the form Ouse\

The analogy then is something of this kind. It is clear that the soft Isc
or Uisc ^ was hardened into Exe and Axe, when the English named their
Exeter and Axminster, and therefore not improbable that the soft Ouse,
or some word with a sound between the Usk, on the west, and the Ouse
on the east ^, may have been hardened for similar reasons into Ox when
they named their Oxford ; while in the latter case there would be the addi-
tional reason of the natural substitute of a knowm word, with a meaning
which was in accordance with its use, instead of an unknown one *.

It is difficult, perhaps, to determine whether the river-name had already
changed from a soft form to a harder form resembling more nearly Ox
before it was combined with ford, or whether, after it was combined with
ford, the Osan became by degrees Oxen ; and in bringing analogies to bear
we are met with historical difficulties ^ Few names of places now remain
beginning with Ouse or Ose. We have beside Osen-ey Abbey, which will
be referred to later on, the well-known instance of the two Ouse-burns
in Yorkshire, which have retained their name, and Ouse-thorpe and Ouse-
fleet in the same county, which probably derive their names from the river.

Atylaiv SevTfpa ae^aarTj. Ptolemj^ it may be added, also gives as one of the cities
of the Dobuni, together with Venta (i. e. Winchester) and the Hot Springs (i. e.
Bath\ "Iaxa>^i^, which may perhaps be identified with Il-chester in Somerset, and
which thus seems also to contain the root word for river.

^ The name Witsa is given in the Anglo-Saxon Chronicles under the year 905 to
the East Anglian Ouse. And the name Usa under the year loio to the same,
with special reference to the upper part near to Bedford. Both Florence of
Worcester and Simeon of Durham employ the term Usa for that river as well as
for the Yorkshire Ouse.

^ The discussion of the root form of the Celtic word for water would be beside
the purpose of these remarks, though Us or Use may be noted as common to both]
Usk and Ouse, and in all probability the Isca of the Romans was a latinization of!
Uisk rather than of Ise. Further, it is possible that Esk was intermediate betweeni
Uisk or Isk, though it must be confessed that no trace of it can be produced in the
south of Britain. The numerous cases in which yEsc appears in the southern
districts are all more readily traceable to the Teutonic for the mountain-ash tree.

^ The name of Ock for the river will be considered later on. See p. 363.

* ' Teutonic nations inhabiting a county with Celtic names have unconsciously]
endeavoured to twist those names into a form in which they would be susceptible
of explanation from Teutonic sources. The instances are innumerable.' Isaac
Taylor's IVoi-ds and Places, second ed., 1S6,;, p. 466; third ed. revised, p. 265.

^ The classical writers do not help us much. Besides the river Isca, Ptolemy'
gives Ou^eAa as a town of the Dumnonii, and just before Tamare and Isca ; while
the geographer of Ravenna gives Uxela and an Uxelis. Possibly the former is
represented by Exeter, and Lostwithiel, on the Fowey, in Cornwall, occupies the
site of the latter, but it seems to show a hardening of the sound of the river
word before the advent of the Saxon. Further, the latter geographer has given
Axium in his list of rivers, but as he did not compile his geography till the
middle of the seventh century, he may have had opportunities of hearing of a later'
nomenclature. At the same time he keeps Isca, intending it for a different river.



APPENDIX B. 355

The names of Ous-den, Suffolk ; Ous-ton, Durham and Leicester and
Northumberland; and Ous-by, Cumberland, may have nothing to do with
the river at all.

In Domesday we find only one or two names which have the prefix Ous,
namely an Ous-torp in Yorkshire, and another in Lincolnshire, an Ouste-
wic (now spelt Owst-wick), and an Oustre-feld in Yorkshire (which does
not appear to have survived) ; also an O use-ton in Bedfordshire, but this
is elsewhere in Domesday spelt Houstone, and is now Houghton Con-
quest. It is doubtful therefore if any of these throw light on the survival
of the river-name Ouse ; but while possibly under other forms the name
has survived, the consideration of the cases would only lead for the most
part into a field of guesswork ^

When, however, we go back to documents before the Conquest, here
and there traces are found of the Ouse, e. g. in Osan-ige and Osan-lea. To
weigh the bearing of the former of these some historical considerations are
necessary to be taken into account, because at first sight it appears to give
certain evidence of the change of Osan-ige into Oxn-ey, and would therefore
distinctly support the change of Osanford into Oxanford.

A copy of the will of JClfric, Archbishop of Canterbury {ob. 1005) is
preserved in the Abingdon series, and judging from internal evidence the
Abingdon chronicler has transcribed the original accurately, and he has
added to it a Latin translation. By his will the Archbishop bequeaths
certain lands to Abingdon but more to S. Alban's, and amongst the latter
he gives land at Great Tew {apud Tiuuan) and Osan-ig, with land in London
which he had purchased '^. According to the Gesta Abbatum Monasterii S,
Alban'i, the early portion of which is supposed to have been compiled by
Matthew Paris himself, a record is introduced of a certain abbot ^Ifric,
who is made to succeed Leofric his brother, who died in 1006. Here, how-
ever, there is abundant reason to suppose that Matthew Paris has confused
the order of the two brothers, and that when the S. Alban'S chronicler
narrates the good deeds of Abbot jElfric, he is practically relating what
Archbishop ^Ifric had done for the abbey, and amongst them he says that
be purchased from the king for a thousand marks Oxon-age and Eadulfin-
tona, which had been put in pledge ^, Here we have mention of land which

^ In the Domesday of Kent there is an Os-pringes, vi^hich name still survives near
to Faversham. And in that of Yorkshire there is an Os-princ, which is now called
Ox-pring (near the source of the Don). This apparently provides an instance of
the change of the Os into Ox.

^ Chron. Mon. Ab., Rolls Series, i. p. 418.

' Gesta Abbatum Monasterii S. Albaiii, Rolls Series, 1867, vol. i. p. 33. See
also the note by Mr. Riley, p. 30, in which he shows that this .^Ifric was the tenth
ibbot and was made archbishop, and that Leofric his brother was the ele\enth.
Further it may be remarked that the bequests of the will and the benefactions
recorded agree in other particulars; e.g. as regards Eadulfinton, the following
appears in the will, in the Latin translation : ' Et hoc apud ipsum suum dominum
srat interveniens ut concederet loco Sancti Albani terram apud Cingesbiri et ipse in
:ommutatione reciperet Eadulfingham.' Further than this, the S. Alban Chronicles
refer to the gifts, both of Tew and Kingsbury. Other points show that the lands
referred to in the different documents are identical.



^^6 THE EARLY HISTORY OF OXFORD.

in the tenth century, as shown by the original will of the Archbishop of
Canterbury, had still the name of Osan-ige ; when, however, the thirteenth
century chronicler puts together his notes he spells the name Oxon-age.
But later still, and throughout the S. Alban's Registers, it is spelt Ox-ey\
and on the Ordnance map Ox-hey"-.

But concurrently with this we find in a volume of thirteenth century
transcripts of documents relating to S. Alban, that of a charter of Ethel-
red, dated 1007^, mentioning certain lands then granted or confirmed
by the King, and amongst them Oxan-gehage, and from the general con-
tents of the charter there is no reason to doubt that it refers to the same
land mentioned in the will of Archbishop iElfric. So that we are met
apparently by the difficulty of two diff"erent names for the ground at the
same time, supposing of course that the transcribers have each followed the
original copy. For it is not simply a change of a few letters. The Osan-
ige of the will is the ' Ouse island.' The Oxan-gehcege of the charter is the
' Ox-paddock.' The form ending in ige or ey or island is followed through
all the medieval documents, as Ox-ey seems to be the name by which their
manor was known to the monks at S. Alban's. The Ox-hey of the map,
however, seems to reflect the hage of the charter.

The argument, therefore, rests upon the priority of the name in the
will of Archbishop jElfric, and on the assumption of it being rightly copied.
It is another question whether the name was derived from the name Ouse
having been once applied to the stream which now bears the name of Colne,
and which is the next important tributary to the Thames, after the Thame
on its northern bank : but it may be added that on this same Colne JJx-
bridge is situated, the origin of which is open to question, since no early
documents have been observed in which the name occurs.

On the whole, then, while it cannot be proved that Osan-ige was the
only name of the place in the tenth and eleventh century, it does appear the
place had that name. How far Oxan-gehag was a contemporary name, or
how far it is possibly due to the gradual change going forward from the
' Ouse island ' to ' Ox paddock,' there are not sufficient data on which to
arrive at a definite conclusion : but in either case we have a precedent for
an interchange of Ouse and Ox.

With respect to Osan-lea *, though the charter which grants the land to
Abingdon Abbey in the year 984 has the boundaries of the land attached,

1 In one place in the Register (p. 4 76), under Abbot Rogers' additions to the
monastery 1260-90, it is spelt Ok-ey, i.e. 'villis de Crokeley, Okey, et Mykelfield.'

- The identification of the district to the south of Watford and on the left bank of
the river Colne, which is marked in the Ordnance Survey in one or two places Oxhey,
is not certain, but no trace of any other is found. It would appear to have been ,
broken up into more than one manor. In Abbot Rogers' time (1260-90) we find |
Oxey Walrand, and in Abbot Thomas' time (1349-96) we find Oxey Richard. I

3 This will be found printed in Kemble's C. D. vol. vi. No. 1304, p. 158. ,

* It consists of a piece of land of two hydes, 'ubi vulgus relatio dicitur set
Osan-lea^ given to Abingdon in 984 {Chron. Mon. Ab., i. p 393). It may perhaps
best be referred to one of the two manors of Ose-lei mentioned in the Domesday ^
Survey, fol. 51a, col. i, in Hampshire. The Yorkshire Domesday has an Ose-h ;
(fol. 316 a, col. 2), but this is less probable. The modern names of neither of these 1
have been identified. ;



APPENDIX B. 357

it has been found impossible to identify it satisfactorily. The abbey does
not seem to have retained it long or it would have appeared again in their
records, but no trace of any sale or exchange, or even the slightest reference
to it, is found beyond this single document, so that any argument to be
derived from it would be valueless. Still, it exists as an early example of
the name Ouse being retained to that date.

These solitary instances rather go to show how much the name Ouse
liad been superseded in all compounds in which the name entered. There
are, moreover, only two charters in which the Use appears named as the
river bounding the properties granted; unfortunately they do not admit of
the site of the land referred to being with certainty identified^.

Next it may be objected that there is no evidence that the river at
Oxford ever bore the name of Ouse ; and further that this river from its
first appearance in history has borne the name of Thames. To meet these
objections it will be necessary to enter into one or two considerations
which may at first sight appear to be somewhat of a digression ; but it
must be borne in mind that the nature of the discussion is a peculiar one,
inasmuch as we are considering historical facts, relating to the district at a
date before we have any historical records of that district ; hence the case
must inevitably rest upon the evidence which can be obtained by, as it
were, reflected light — reflected, that is, from events and circumstances
belonging to a much later period.

The objection arising from the known name of the river being theThames,
and therefore that the name Ouse is excluded, is met at once by the con-
sideration that the name Thames or Tem-ese is composed of two words,
the latter, as has been noticed, being a dialectic form of the river-names
amongst which also the Ouse is included. The form in which the name
Thames first comes before us does not mihtate against this view. In the
two places ^ where it is mentioned by Caesar, in his Commentaries, written
B. c. 55, it is always spelt in the MSS. as Tam-esis. In the annals of
Tacitus '\ written circa A. D. 80, it is written Tam-esa. In the Geography
of Ptolemy {c. A. D. 120) several rivers are given, but unfortunately the
arrangement he has followed involves his omission of the Thames, though

' In the grant of King ^Ethelred to a certain Elf here in 979 of land at Ollan-ege,
it would perhaps appear that Olney in Bucks must be meant, since it lies on the
banks of the Ouse. Amongst the boundaries occur 'Andlang broces inon Use;
andlang Use on Wilinford ; of tham forde andlang Use to Kekan were ; of Kekan
were andlang Use on Caluwan war' (Kemble's Codex Diplomatiais, vol. ill. No. 621).
But no traces of these names now exist. In the case of Niwantun, there are so
many Newingtons it is impossible to say whether there is a river near the place
still bearing the name or whether the word is meant for a river at all. The
boundaries run, ' And than to Use stccthe on Ealferthes hlaew ' (Kemble, C. D.,
vol. V. No. 1 1 14). The form Usan-mere and Us-mere are also found, which
probably refer to the river.

^ Caesar de Bella Gallico. 'Cujus fines a maritimis civitatibus flumen dividit .
quod adpellatur Tamesis ' (Lib. v. § 11). 'Caesar cognito consilio eorum ad
flumen Tamesin in fines Cassivellauni exercitum duxit ' i^ibid. § 18).

' Taciii Annales. ' Visamque speciem in aestuario Tamesae subversae coloniae '
(Lib. xiv. cap. 32).



358 THE EARLY HISTORY OF OXFORD.

he names amongst those on the southern shore of Britain both the Tamar
and the Isc or Exe ^.

In the third century we have Dion Cassius copying Tacitus and spelling
the word lam-esa. But few other classical writers name the Thames^, and
with these, as well as with our own early writers, we find much the same
spelling adopted. We cannot trust, perhaps, the MS. copies of Gildas and
Nennius, each of whom once mention the Thames ; the best MSS., how-
ever, give respectively Tham-esis and Tam-isia. In Beda we have in one
passage Tam-esis, but in all others the adjectival form of Tamensis
fluvius. Probably, however, the earliest and the first actual English form
is that given by the Anglo-Saxon Chronicle under the account of Caesar's
invasion, where it is spelt Taem-ese. Two of the later copies make this
Temese, and throughout the rest of the earliest copy of the Chronicle,
and in most of the others wherever the word occurs (and it occurs about
twenty times), it is written lem-ese. In this it would appear that the two
component parts of the word as understood by the people of the country
are clearly set forth, i, e, lem and ese.

If we turn to the Charters we find the more common form of the Tem-ese
sometimes Tam-ese, rarely Tem-is, and its Latin representative Tam-isia.
The medieval historians adopt mainly Tamesis, and very frequently the
adjectival form of Tamensis.

The first word of the compound, namely Tam, is found preserved in
other rivers throughout the country, i. e. the Tamar ^, the Teme, the
Thame, &c., and they have left their names in towns such as Tamborn
and Tamworth in Staffordshire, a Tamworth in Warwickshire, and a
Tamerton both in Devonshire and in Cornwall. There has been some
discussion about the meaning of the word, but probably 'wide spreading'
is the simplest as well as the most accurate interpretation, the name being
derived from the large expanse of meadow land which is laid under water
in flood time, when the river overflows its banks.

The suffix of eje, however, is the most important. Neither ese nor ije are
known to exist as distinct names of rivers in this country, but, as already
pointed out, they must belong to the same class as the Uisk and the Ouse,
and if we go across the Channel we find the Oise, which gives its name to
two Departments, and which must be a member of the same family. Further,
the Isca of the Antonine Itinerary and the "laaKa of Ptolemy, as already
pointed out, prove the existence of a form approaching Ise.

The direct evidence now for the Thames at Oxford having been called
the Ouse lies in the circumstance that one of the islands, which may well
have received its name as early as Oxford itself, became, in the year 1129,

1 Claudii Ptoloniaci Geographia, Lib. ii. cap. 3. Tafiapov nor. eK^oKav. —
'laoLKa TTOT. kK^o\av.

* Pauhis Orosius, writing in 417 of Caesar's invasion, according to the MSS.,
has Thamesis as the name of a river, but^the anonymous Geographer of Ravenna,
who wrote ciixa 650, puts Tamese amongst his list of cities next after London.

^ The Tamar is found noted as a river by Ptolemy (see above, note i). He
also amongst the cities of the Dumnonia gives the names lajxapr] and "loKa. The
anonymous Geographer of Ravenna {circa a. d. 640) gives amongst cities Tamaris. *
Possibly Tamer-ton occupies the site.



APPENDIX B. 359

the site of an important monastery, and that form of the name was at once
stereotyped, so to speak, in the Abbey of Osan-eia, that is the island of the
Ouse^, so that the form has been retained to the present day, without
having been corrupted into any other. The variations of the name itself
amongst the charters and chronicles are very limited, seldom departing
from Osan or Osen for the affix, and when written in full either ea or eta or
eye'-, for the suffix. This is not only shown throughout the large number
of charters which are found in the cartulary of the abbey itself, but from
the frequent reference to Oseney in other cartularies, as well as from the
entries in Rolls of various kinds. One exception has been observed, namely
in a single charter in the Bodleian Collection^, in which the 'Abbot of
Oxen-ea ' is named ; the charter is not dated, but internal evidence shows
it to be about the middle of the thirteenth century *. No other charter in
the same collection (and there are several with Oseney mentioned in one way
or another) has this form, nor has it been observed in others ; so that a
late solitary exception amongst perhaps a thousand instances, dating from
the early part of the twelfth century onwards, must be attributed to the
error of the scribe rather than any chance survival of an earlier form of the
name. That it might represent what the scribe had heard the place some-
times called in his time is possible, but then this would rather point to
that tendency, already described, of the substitution of a known name for
an unknown ; in other words the occasional corruption of Ousen-eye into
Oxen-eye would rather illustrate the growth of Ox-ford from Ousen-ford.
That it has even been called Oxeny in modern times would have been
doubted, were it not that on the Ordnance map of 1 8 30 Oxny Mill appeared^.
The evidence taken as a whole, therefore, points to the fact that an '^ eye^
or island in the river at this point bore the name of Ousen-eye from the
earliest time when it comes into history ; and as it is contrary to the order
of the corruption of names that a known English word should be cor-
rupted into an unknown of the same length, it must be assumed that
the name Ousen-eye was given at the time when and for the sole reason
because the river there was called the Ouse.

Another singular piece of evidence may here be adduced which seems to
show the connection or rather the identity of the Ese and the Ouse, that is
to say, the same river may have been called sometimes one, sometimes the
other. There is an important river now bearing the name of Ouse, which

1 Isaac Taylor, however, in Words and Places, finds in the n in Ose-n-ey not
the Saxon genitive, but ' probably a relic of the Celtic innis or island ' (ed. 1 865,
p. 204).

^ In one case, in a Close Roll of King John's reign (anno 15, memb. 9), the form
Osen-heye has been observed.

^ For this important example the writer is indebted to E. B. Nicholson, M.A.,
Bodley's Librarian.

* The charter runs, 'Sciant omnes gentes quod ego Aeliz Ghemio quondam
uxor Henrici de Tomelee defuncto dedi, etc. ij solidos redditus quos abbas
Oxenee reddit ibi annuatim.' No. 439, Original Charters, Oxfordshire : described
in Turner's Bodleian Charters, 1878, p. 372.

' In later editions of the map, since the railways were inserted, it appears to
have been omitted. In the 25-inch scale map it is spelt Osney.



360 THE EARLY HISTORY OF OXFORD.

rises near the southern borders of Northamptonshire. It flows past
Buckingham, Stony-Stratford (where the Watling Street crosses the
stream), and Olney ; but before it reaches Bedford it is joined by a tributary
called the Ousel, which rises at the northern extremity of the Chiltern
Hills, near to Dunstable \ The main stream, after passing Bedford, flows
by a town named Tempsford ^, and then passes Huntingdon and St. Ives,
soon after which it receives the waters of the Cam, and then passes Ely,
and draining the great fen country of Cambridgeshire and Norfolk flows
into the Wash. In this part of England it must have appeared as important
as the Tam-ese itself. It is therefore at first sight a very singular fact that
a river which must have once borne the name of Ouse^ and bears it
still, should ever have boi-ne at any time, with any settlers, the name of the
Tam-ese. And yet it was so, for at the place where the river was forded a
town grew up which bore the name of Tam-esan-ford, i. e. the ford of
the Tam-ese.

Under the year 921, in describing the ravages of the Danish army about
Bernwood and Aylesbury, the Anglo-Saxon Chronicle narrates : —

* At the same time the army from Huntingdon and from the East

Angles, went and wrought the work at Teem-ese-ford, and inhabited it,

and built, and forsook the other at Huntingdon ^.'

Nearly a century later on — that is, under the year loio — and in describing