Origin of the Anglo-Saxon Race/Chapter 23

by Thomas William Shore

CHAPTER XXIII.

SETTLEMENTS ON THE WELSH BORDER.

AT an early time in the Saxon period the district which is now Gloucestershire became a frontier country. It was opened to settlement on the east of the Severn by the victory of Ceawlin, King of Wessex, at Deorham in 577. The Severn then became the boundary between the Britons and Saxons, and the county was down to a late period considered to be within the Marches of Wales. The Gloucestershire country east of the Severn, which was originally part of Wessex, became later on separated from it under the rule of Ceolric of the West Saxon royal house, and was subsequently absorbed by Mercia. This is of interest in pointing to the direction from which this county probably received its earliest Saxon settlers. The early administration of this district appears to have been connected with Gloucester, Berkeley, Tewkesbury, and Cirencester. There was an extensive administrative area attached to Tewkesbury as late as the Norman survey. The Berkeley administrative area was also large, and was known for many centuries as Berkeley-herness. This name appears to be Scandinavian, and, like those of Inverness in Scotland, Agremundreness in Lancashire, and Holderness, the Berkeley district as a separate area may have had a Scandinavian origin.

In Gloucestershire, as in the northern counties, the evidence of earlier Scandinavian settlers is much mixed with that of the later, so that it is not possible in some localities to distinguish the earlier from the later.

The evidence of Northern settlers, whether of earlier or later date, is remarkable. Near Bristol is a place called Yate, the Geate mentioned in several Saxon charters. Another old name, probably denoting the settlement of a Goth, is Mangotsfield; Hacananhamme, or Hacon’s ham, the old name^[1] for Hanham, near Bristol, is clearly Scandinavian. In Gloucestershire, and close to it along the Wye, there were small areas called shires, corresponding to hundreds similar to the shires in the northern counties, and to the shires of ancient Norway. There are old records relating to Blakeborneshire and Pignocshire, near the Severn and the Wye. Huntishamshire was the name for a detached part of Monmouthshire, near Welsh Bicknor. In the south of the county, also, is an old hamlet called Kendalshire. The name Scir-mere occurs in a Saxon charter, and the modern name Shirehampton, nee Bristol, may be a survival of one of these old names.

The name Berkelai-erness, as already mentioned clearly corresponds to those of Holderness and Agremundreness, both of which received Northmen among their colonists. The termination -ærnes is common among the place-names of Scandinavia. The tidal bore in the Severn at the time Camden wrote still retained its Scandinavian name Hygre, derived from the Norse mythological name Oegre, the Neptune of Northern tribes. The Scandinavian name Brostorp is a Domesday name near Gloucester, south of which place are also Brookthorp and Calthorp.

The dialect of the vale of Berkeley differs both in words and pronunciation from that of the vale of Gloucester, higher up the river.^[2] As already noted in relation to Somersetshire, the Scandian name Holm appears in the names Great Holm and Flat Holm for islands at the mouth of the Severn. It occurs also in Holm Lacy, near Hereford. Some remarkable Scandinavian names are found along the lower course of the Severn. Sanagar, anciently written Sevenhangar, is that of one of the old tythings of Berkeley, and Saul, also near the river, reminds us of the Saul district and the Saulings, whose name is mentioned in a runic inscription at Glavendrup in Scandinavia.^[3] The forest district of Dean between the Severn and the Wye was, apparently, named after Dene, for Dane, and not den, a wood. Giraldus Cambrensis, writing in the twelfth century, tells us that the Dean Forest was known in his time as Danubia and Danica Sylvia, or Danes’ Wood.^[4] The name Danube for the country of the Danes is an old one. Asser, in his ‘Life of Alfred,’ says that in the year 866 a large fleet of pagans came to Britain from the Danube. The old name Dene for this forest district appears thus to be that of Dene, the Danish name, and it is still called Dane in the local pronunciation. The language of the ancient Northmen has survived to the present day in the name Aust, anciently Austrecliue,^[5] or Aust cliff, on the east side of an ancient ferry across the Severn, near Bristol, austr being the Old Northern word for east.^[6] Mona is a variation of the name of the stream called Monow, which joins the Wye at Monmouth, and Mona is the latinised form of the name of the Danish island Moen, or Mön.

Ethelweard tells us in his Chronicle that in 877 the Danes made a settlement of some kind in Gloucester. The custom of borough-English still survives there, as it did at Stamford, Nottingham, Derby, and Leicester, all of which were Danish towns, and we may reasonably connect the custom at Gloucester with some of the so-called Danes who may have settled there. The custom was probably brought by some allies of the real Danes, perhaps people of the Wendish or a mixed race. The custom of the real Danes of Old Denmark was that of partible inheritance.

Gloucestershire had a custom that resembled one of those of Kent—viz., that under which the lands and tenements of condemned felons were not forfeited. They were only held by the Crown for a year and a day. In this we may see a resemblance both to the custom of Kent and that of the Archenfeld part of Herefordshire.

The settlements in the lower parts of the valleys of the Severn and the Wye appear to have been effected by direct maritime migrations. The ships of the period could ascend these rivers by aid of the strong tide which flows up them to the neighbourhood of Gloucester and Monmouth. Goths or Kentish colonists on the Wye have not only left a trace of their name in that of Goderich, now Goodrich, but also in some of their customs in the district known as Ircingafeld^[7] or Archenfeld. It comprised the south part of Herefordshire, having the Wye on the east and Monmouthshire on the south. Some remarkable old Kentish place-names can be traced within or near it, such as Kentchurch, Kenchester, Kentyshburcote,^[8] and Kenthles.^[9] These names, together with the customs which prevailed, show that the Herefordshire province of Archenfeld must have received Kentish people among its Gothic or Jutish settlers, who had no doubt inferior Welsh tenants under them. The local customs of Archenfeld closely resembled those of Kent. That of partible inheritance, of the same nature as Kentish gavelkind, survived in the district until it was abolished in the reign of Henry VIII. This Kentish custom differed from the partible custom that prevailed in Wales in three essential particulars, which will bear repetition: (1) By the Kentish custom in Archenfeld only legitimate sons inherited the paternal estate. By the Welsh custom all sons, legitimate or otherwise, had their shares, or in early centuries fought for them. Giraldus, writing in the twelfth century, tells us of the contention of legitimate and illegitimate sons for shares of the paternal estate. (2) By the custom of Archenfeld, like that of Kent, daughters inherited if there were no sons. Under the Welsh custom they did not. (3) By the custom of Archenfeld, like that of Kent, widows were entitled to their dower of half their husbands’ customary estate. Under the Welsh custom they had no dower.

The resemblances between the other local customs are also remarkable. In Kent, if a tenant in gavelkind was convicted of crime and executed, his land was not forfeited, but went to his heirs. This was known as ‘the father to the bough, the son to the plough’ custom,^[10] and was a rare privilege,^[11] which the people of Archenfeld also had. In Kent, a tenant in gavelkind had the power of bequeathing his land to whom he pleased, and the people of Archenfeld had a similar privilege in respect to land they acquired.

The most remarkable of these parallel customs is, however, that under which the men of Kent claimed as their immemorial right the privilege in war of being marshalled in front of the King’s army, a claim that was recognised.^[12] The men of Archenfeld claimed and had allowed to them the same honourable distinction.^[13] These remarkable coincidences clearly indicate a Kentish colony.

This district of Herefordshire appears to have been in any case occupied by Teutonic settlers at an early period, and to have become an outlying part of Mercia by the end of the seventh century. Ceolred, King of Mercia, dated a charter ‘in loco Arcencale,’ probably only a variation of the name, early in the eighth century.

The Scandinavian evidence already mentioned points to a later settlement between the Severn and the Wye, and also in the north of Monmouthshire. This country and that near the west of Herefordshire was part of the district of the Dunsetas, where English settlers of some kind lived side by side with the Wealas or Welsh. In Ethelred’s ordinance relating to the Dunsetas^[14] provision is made for diffusing among them a knowledge of the laws they were required to obey, and it is expressly stated that twelve lahmen shall explain the law to both the Wealas and the English, of whom six shall be English and six Welsh. The significance of this ordinance is in the legal terms used^[15] lahcop, Old Norse lögkaup, and witword, Old Norse vitorth. The term lahmen is also Danish, and is mentioned in Domesday Book in connection with the administration of the Danish towns, such as Stamford. The names lawrightmen and lawmen survived in Shetland until comparatively modern times.^[16] There is also a reference to the twelve lahmen in the ‘Senatus consultum de Monticolis Walliæ,’^[17] who were, apparently, the successors of those appointed for the Dunsetas a century earlier. If the English people among the Dunsetas had not been of Danish or Northern descent, Norrena or Danish names for legal officials and legal terms would not have been used in this ordinance. Sweden and Gothland in olden time were the land of lagmen or lahmen, for the whole territory was a confederation of commonwealths, each with its assembly of freemen, law-speaker and laws.^[18]

From the evidence relating to Archenfeld there can be little doubt of an early settlement of Kentish colonists or Goths in that district, as there was, perhaps, in other parts of the same county, and a later settlement of Northmen. The only record of any political connection between Kent and Herefordshire occurs in the seventh century, when Merewald, viceroy of the Hecanas, or tribal people of that county, and brother of Wulfhere, King of Mercia, married Eormenbeorh, a princess of Kent. She was a granddaughter of King Æthelbert, and a cousin of Eormengild, who married King Wulfhere. Between the royal houses of Kent and Mercia there was by these marriages a double alliance. Merewald was also called ealdorman of the West Angles. In the eighth century we read of Arcencale as apparently part of Mercia, and by that time it had perhaps already received its Kentish or Gothic settlement, of which Goderich became the administrative centre. It is probable also that before the time of Ethelred II., King of Wessex, there had been a further settlement of Danes or Northmen along this Welsh border, seeing that officials with old Danish titles were appointed to explain the laws to the Dunsetas.

One of the proofs of Scandinavian settlements in the border counties is the hope place-names. Among the names on the coasts of Scotland and in the parts occupied by Scandinavians in that country are a large number of hope names. There were sea-shelters so named by them, such as Long Hope, Kirk Hope, Pan Hope, and St. Margaret’s Hope, in the Firth of Forth, another in the Orkneys, and Gray Hope in Aberdeen Bay. The Norse settlers in the south of Scotland also gave the name hope to inland places which were shelters between hills. There are sixty hopes in the counties of Peebles and Selkirk alone, and many more in Roxburghshire and the Cheviot country.^[19] The derivation from hóp, Icelandic, an inlet of water, is clear for the sea hopes, and in the sense of land havens in exposed hilly regions for the inland places so named. The termination -hope is often pronounced -op and -up. The place-names along the Welsh borderland show some remarkable examples of this kind—i.e., places with names ending in -op and -hope. In the east of Radnorshire we find old places named Cascop, Augop, and Hope; in Shropshire, Hope Bagot, Hope Bowdler, Hope Hall, and Hope Sey; in Herefordshire and along the Gloucestershire border, Hopend, Faunhope, Woolhope, Hope, Hope-Mansel, Longhope, Arcop or Orcop, Brinsop, Seller’s Hope, and the Domesday name Gadreshope.

Wigmore, Wormsley, and Ross appear to be names of Northern origin. The old district shire names already referred to are also remarkable. The Scandian termination -ore appears in the names of English and Welsh Bicknor, Yasor, Eastnor, and Radnor.

The Herefordshire Domesday hundred names include those of Radelau, Thornlau, and Wermlau, which appear to be of Northern origin, and at Marden in this county the old Norse custom survived by which in default of sons the eldest daughter succeeded to the whole inheritance.^[20]

The Teutonic colonies on the coast of South Wales have been commonly ascribed to the Flemings settling there in the late Norman period. The dialect of Gower and Pembrokeshire, which resembles the West Saxon, shows, however, no trace of Flemish influence. A. J. Ellis, who investigated this subject, says that at most there could only have been a subordinate Flemish element, which soon lost all traces of its original but slightly different dialect, while the principal elements must have been Saxon, as in Gower and the Irish baronies of Bargy and Forth, in the south-east corner of Ireland.^[21] A Flemish settlement in South Wales is historical, but the loss of all linguistic traces of it shows that the descendants of these settlers were absorbed among the much larger population of Saxon and Scandinavian descent previously located there. This view is supported, first, by the place-names which are of the West Saxon and Scandinavian types; and, secondly, by the customs of a large number of manors in Glamorganshire, which are different from the Welsh, and bear a close resemblance to those in west Somerset, to which locality the dialect also points. There must have been a connection between the settlers on both sides of the Channel, as the dialect, customs, and general character of the old names show.

There is also evidence which shows that the coast of Wales and its border near the sea was occupied by Anglian settlers at an earlier period than that of the main settlements by Northmen, and this may be summarised as follows: The topographical name Angle survives on the coast, and can be traced also in old records on the north-east border of Wales. There are Anglesea on the north-west, Angle and Angle Bay in Pembroke Harbour, and Pen Anglas, a promontory west of Fishguard Bay. In the Patent Rolls 9 Edward I. and other documents we read of the cantred of Ross and Englefeld, in or very near the county of Chester. It may be considered certain that these Angle place-names were not given to the districts to which they refer by the Welsh. Their name for Angles, Saxons, and Jutes alike was Saxons, similar to the popular Irish name for the English at the present time. These old place-names must have originated at a time when Angle was in use as a distinctive name for people who migrated from England or for settlers from the North of Europe. During the Viking period such settlers would be known as Northmen and Danes. It is not at all probable that the Angles of Northumbria or parts of Mercia formed new settlements on the Welsh coast while the Danes were establishing others on their own, and it must be remembered that after the Danish period the name Angle or Engle passed out of use, and the English name became solely used. It is difficult, therefore, to avoid the conclusion that these Angle names on the Welsh coast must have arisen some time before the earliest Danish inroads.

The probable Anglo-Saxon occupation of parts of the coast of South Wales before the period of settlements by the Northmen is confirmed by the prevailing dialect. Ellis says: ‘The south-west of Pembrokeshire, or two peninsulas at the south-west, form an old English colony.’^[22] He points out that the character of the dialect of this part of Pembrokeshire is decidedly southern, having such examples as dr for thr in three, through, and threaten; having v for f in fair, farm, fast, feel, fiddle, fox, flail, from, and furrow; having z for s in say, self, seven, sick, six, soon, and Sunday; while s remains with less regularity in sad, sand, saw, so, and sweet. He likewise says: ‘The peninsula of Gower is also a very old English colony, consisting of seventeen English parishes.’ He remarks that the reverted r is inferred from the word drou for through, and that there is an occasional use of z as an initial sound for s, and un as an unaccented word for him. These examples are distinctly southern English, but the dialect in Gower seems to have much worn out. With this evidence, side by side with the English place-names, and the prevalence of manorial customs in the vale of Glamorgan identical with those in the vale of Taunton, the supposition that the English characteristics of the people in these parts of South Wales are due to the Flemings entirely breaks down.

One of the most interesting of all the English district names is that anciently given to Pembrokeshire, Anglia Transwalliania, or ‘England beyond Wales.’ That it must have been a very old designation is probable from the surviving Angle place-names in this county, which clearly point to early settlements.

Isaac Taylor says: ‘The existence of a very early Scandinavian settlement in Pembrokeshire is indicated by a dense cluster of local names of the Norse type which surrounds and radiates from the fiords of Milford and Haverford.’^[23] There is other evidence pointing strongly in the same direction, which the same author has mentioned. This refers to the inscriptions known as oghams. The ogham inscriptions which have been found in Wales are about 20. Of these, 17 have been discovered in the counties of Pembroke, Cardigan, Carmarthen, and Glamorgan, 9 out of the 17 having been found in Pembrokeshire. In Devon, 2 ogham inscriptions have been discovered, and 1 in Cornwall. In Ireland they are much more numerous, 155 having been found, but of these, 148 belong to the four counties of Kilkenny, Waterford, Cork, and Kerry—i.e., roughly speaking, they fringe the line of coast which stretches between the two Scandinavian kingdoms of Waterford and Limerick,^[24] thus clearly showing their Scandinavian origin. Oghams are, indeed, a variation of runic writing.

The custom of borough-English is certainly not a relic of Welsh law.^[25] In parts of South Wales it prevailed, with similar privileges to widows as in the vale of Taunton and on so many manors in Sussex. This custom of some manors in Glamorganshire and Pembrokeshire, by which the youngest son succeeded to the whole of the father’s land, must have been introduced by settlers of another race. It prevailed on many lands in Gower;^[26] it was the custom of the manors of Llanbethan,^[27] Merthyr Mawr,^[28] Coity Anglia,^[29] and others. It was also the custom on some of the manors of the Bishop of St. David’s.^[30] The resemblance between this custom as it prevailed at Coity Anglia and the many manors of Taunton Dean in Somersetshire is very close—practically identical. The name Anglia attached to this manorial name is of special significance, for here we find a name with a special Old English designation, having an Old English custom. The custom points to Somersetshire, the dialect of Pembrokeshire and Gower point to the same part of southern England, and there are traditions which indicate this locality as the district whence the Old English colonists of South Wales largely came. Gower is visible in clear weather from the West Somersetshire coast looming in the distance across the Severn Sea. Rhys has drawn attention to a tradition^[31] in connection with the Welsh Arthurian legends, which makes Melwas king or lord of a winterless glass island. This he identifies with Glastonbury in the Æstivo regio, or summer region—i.e., Somersetshire. Another and a different tradition makes Melwas king of Goire, or the peninsula of Gower seen from the Somerset coast. Thus, by mixing the two versions of the myth, the writers of romances came to speak of the kingdom of Melwas as Goire, and of his capital as Bade or Bath. The curious aspect of these traditions is that there may have been a basis for connecting Gower with Somerset; and, as Max Müller says on the growth of myths, there may have been circumstances or words, ‘understood, perhaps, by the grandfather, familiar to the father, but strange to the son, and misunderstood by the grandson.’

As regards the settlement of north-east Gloucestershire, Beddoe has observed the blonde character of the population in the country around Moreton-in-the-Marsh, and considers it evidence of West Saxon colonisation northwards.^[32] He notes that the distribution of colour of hair and eyes in this district resembles that found in other Saxon districts in England, and also in parts of Flanders, Holland, Friesland, and Westphalia, with the same tendency to the conjunction of hazel and dark eyes with lightish hair, rather than of light eyes with dark hair. The head form also he judges to be mostly of the two types found at Bremen, which are also those commonly found in Anglo-Saxon graves. He says that the West Saxons appear to have settled numerously in the Upper Thames districts before they began to interfere with the inhabitants of the valley of the Bristol Avon—i.e., they pushed their settlement northwards at first rather than westwards.

In the same district, near Bourton, in north-east Gloucestershire, the Anglo-Saxon place-names Cwéntan^[33] and Cwénena-broc^[34] occur, referring to Quinton and to a stream which is named as a boundary.

The name Cwénena-broc brings us to a curious difficulty—viz., to determine whether Cwénena is the genitive plural of Cwén, a Fin, or Cwén, a woman. It has been explained as the women’s brook,^[35] but the name Cwéntan, now Quinton, mentioned in a Saxon charter, is in the same locality. There is a well-known story of Adam of Bremen being present at a conversation during which one of the old Scandinavian kings spoke of Quénland, or Quéna-land, the country of the Quéns or Quains. As the stranger’s knowledge of Old Danish was very imperfect, he supposed the king had said Quinna-land, the country of women or amazons. Hence arose the absurd story of the terra feminarum, or amazons’ country, which spread through the whole of Europe, through mistaking the name for that of a woman.’^[36] The name Cwénena-broc must mean either the brook of the Quéns or Fins, as allies of Scandinavia and their descendants, or that of a community of women. Which is the more probable? It is a boundary name, apparently a boundary of Cwéntan, and we must either recognise a settlement of Fins or a settlement of women. During the period when the dialects of many tribal people were being assimilated into one form of speech it is not difficult to suppose that Cwénena may have been written for Cwéna, the usual form of the genitive plural of Cwén, a Fin.

In east Gloucestershire there were also two distinct places called Quénintune at the time of the Domesday Survey—one near Fairford, the Domesday Fareford, and the other in the north-east, apparently the Cwéntan of the Anglo-Saxon period, of which Cwénena-broc was a boundary. It thus appears probable that there were two settlements of Quéns. That there were Scandinavian settlers with whom they probably came as allies, and in whose language Fins were called Quéns, also located on this borderland of Gloucestershire and Oxfordshire, is certain from the old place-names of the district. There are, or were, not less than nine places with the characteristic -thorp or -throp names in this locality. In Domesday Book, Dunetorp, Duchitorp, and Edrope, now Heythrop, are mentioned on the Oxfordshire side. In Gloucestershire there are Addlestrop, Hatherop, Southrop, and Wiiiiamstrop. Tadilsthorpe, the Domesday Tedestrop, and Burdrop, are also old place-names. Among others of Scandinavian origin in the district are Wickenford or Wickhamford; Meon, the Domesday Mene, which may be compared with the Jutish places called Meon in Hampshire; Fareford, Wormington, Guiting, and Sclostre, now Slaughter. Rollright, the Domesday Rollendri, also occurs on the Oxfordshire side of the border, and at this place there is a rude circle of stones of the Scandinavian type. These names, together with that of the Domesday hundred name Salemanesberie, apparently derived from the Salemen or Salings of one of the Danish islands, in which hundred Bourton, Broadwell, and Slaughter were situated, are evidence that there must have been in this district of north-east Gloucestershire many settlers who spoke the old Danish or Norrena language, in which Quén is the name for Fins. Moreover, at Sclostre, now Slaughter, at the time of the Domesday Survey, the rents of two mills were paid in Danish money computation. When King Eadgar promulgated his laws in these words, ‘Let this ordinance be common to all the people, whether English, Danes, or Britons, on every side of my dominions,’ he must have had in mind settlements of Danish-speaking people in the south and west of England, such as this in Gloucestershire, as well as the greater Danish settlement in the northern and eastern counties.

There is evidence, in addition to that of existing place-names, which points to the settlement of some Hunni or Hunsings in the valley of the Worcestershire Avon. There are two Saxon charters relating to grant of land at Hampton, close to Evesham, which in the eighth century bore the name of Huntena-tun, the -tun of the Hunte or Hunsi, the name being mentioned in the genitive plural in both charters—one a grant by Aldred with leave of King Offa, dated 757-775^[37]; the other a grant by King Acgfrid, dated 790.^[38] In a charter of Eadgar, dated 969, relating to land at Witney,^[39] there is a reference to the same settlement in the boundaries, the name ‘huntena weg’ being mentioned—i.e., a road that led to the Huntena district, or Huntena-tun. A few miles east of the Anglo-Saxon Huntena-tun is Church Honeybourne, with its hamlets Cow Honeybourne and Honeybourne Leasows. These surviving names and the reference to the Huntena show that there was a settlement of people who bore that name in this district, and it should be remembered that in the old country of the Hunsings and Frisians there is a river called Hunte, as well as the Hunse.

Reference has already been made to the fair aspect of the people of east Gloucestershire at the present time. The circumstantial evidence of the place-names points to the settlement of tribal people of various blonde races in this district. Among such races are the Fins, concerning whose aspect the proverbial expression ‘as blonde as a Fin’ is in use among the Russians of the parts adjoining Finland at the present day.^[40] The Fins that settled in England must have come as allies of the Danes, and it is interesting to note that by the Roman road east Gloucestershire was in direct communication with Lincolnshire.

One of the peculiarities of the topography of Shropshire and Worcestershire is the considerable number of old names we can trace that apparently denote tribal settlements, as if a number of different people were settled on this borderland in large communities for defensive purposes. Among these, the following are mentioned in the Anglo-Saxon charters^[41]: Wrocensetna and Scrobsetan in Shropshire, and the Tonsetan or Temsetan somewhere west of the Severn. The latter may be the settlers on the river Teme, whose name can still be traced in that of the ancient manor of Tempsiter, which included twenty-three townships of the Honour of Clun, and through which Off as Dyke passes.^[42] The river Clun is the longest tributary of the Teme, the latter name being now applied to the stream only after its junction with the Onny near Ludlow. The Tonsetan or Temsetan appear to have been the settlers on the Welsh border near Clun. Another Worcestershire settlement which is described as a province was that of the Usmére people,^[43] whose name appears to have been lost. In Herefordshire and a part of north Gloucestershire the tribe known as the Magesætas were located. We read of a grant of land at Hay ‘in pago Magesætna’ as late as A.D. 958.^[44] This tribe must have been a large one, and Maisemore near Gloucester may have been its eastern limit. May Hill near Ross, and another May Hill near Monmouth, are probably places where the name survives.

The settlements of Gewissas, by the victories of Ceawlin in the Severn Valley, extended not only over parts of East Gloucestershire, but probably further northwards. Ceawlin’s victories opened the country more or less as far as Shropshire. The earliest colonists into this part of England must have come either up the river or along the Roman roads, the Fosse way from the north-east, the Watling Street from the south-east, or from Wessex by the road from Winchester to Cirencester, and thence by the Fosse way to the north-east of Gloucestershire, and northwards by the Ryknield Street. It was probably about A.D. 583 that the Roman city of Uriconium was destroyed. It was situated where Wroxeter now is, close to the lowest ford across the Severn, south of Shrewsbury, where Watling Street crossed the river. Its remains show its importance, and probably many buildings of the Saxon time in its neighbourhood were constructed from its ruins. In the Severn Valley there is historical evidence of the settlement of West Saxons, and that about 590 an independent State of Gewissas was formed in Gloucestershire under Ceolric, a nephew of Ceawlin.^[45] The dialect also points to its settlers having largely come from Wessex. Ellis groups it with Wilts, Berks, and parts of Hants and Dorset, as districts having much in common.^[46]

Anglian settlers from Mercia or others who had a knowledge of runic letters appear to have reached the south-east of Shropshire by the end of the sixth century, for a runic inscription discovered at Cleobury Mortimer has been assigned by Stephens to that period.^[47]

It may have been from the circumstance of the ruined condition of the Roman city Uriconium that the Saxon colonists near it got their name of Wrocensetna, as Camden suggested. It may, perhaps, have arisen partly from the settlers having made a quarry of the ruins. Almost all the stones in the walls of Roman Uriconium were removed, as well as the ruins of its buildings, and from the wrecked city there can be no doubt many a house, or even in later centuries a church, was partly built, as may be traced around Silchester, where the destruction was less complete. The Wrocensetna have either left their name in that of Wroxeter, the village on the site of Uriconium, or derived their name from it, the wrecked ceaster. The name survives also in those of Wrockwardine and the Wrekin. The pagus or province of the Wrocensetna is mentioned in a charter of Burgred, King of Mercia in 855,^[48] and in one of Eadgar, dated 963.^[49] The survival of the word ‘wrocen’ or ‘wrekin,’ as probably a reference in Saxon nomenclature to the ruins of a Roman city, is unique among English topographical names.

The ancient name Ombersley in Worcestershire, whose early settlers are called the Ombersetena, is as old as the Saxon period.^[50] These people, whose name has come down to us in the genitive plural, are probably the same as the Ymbras or Ambrones—i.e., the tribe of Old Saxons south of the Humber. This colony of them in Worcestershire was probably a migration from their district on the Amber River in Derbyshire, from Nottinghamshire or Lincolnshire, along the Roman roads that passed from Chesterfield through Lichfield into Worcestershire. They apparently gave their new settlement the same name, which some of the tribe had brought from the Ambra River in Old Saxony.

In Shropshire an interesting peculiarity has been observed in the country dialect. This, according to Prince L. L. Bonaparte, is the verb plural ending in n, as ‘we aren’ for ‘we are,’ and also the form ‘we bin’ for ‘we are.’ This he points out as an interesting instance of the shading of the southern dialects, in which ‘I be’ and ‘thou bist’ are common, into the north-western.^[51] That some settlers in Shropshire came up the river is probable from the dialect and from some of the customs. Borough-English, which still survives at Gloucester, prevailed in the English part of Shrewsbury.^[52] In this county also there were, at the close of the Saxon period, tenants called coscets, few in number; but as coscets are peculiar to Wiltshire, these may have been descendants of Gewissas who had migrated.

Along the border counties of Wales there was necessarily going on during the Anglo-Saxon period some racial fusion between the tribal people respectively of the Teutonic and Welsh races. As the Welsh were driven westward from the Midland counties, their agricultural system of isolated homesteads appears to have been commonly adopted. Villages of collected homesteads, like those between the Elbe and the Weser, or east of the Elbe, and such as are found in Northamptonshire and the adjacent counties, are comparatively rare along the Welsh border. Giraldus tells us that in the twelfth century the houses of the Welsh tribesmen were not built either in towns or villages. Like other pastoral people, they had two sets of homesteads, feeding their herds in summer on the higher ranges of the hills and in winter in the valleys. The Old English settlers along the border counties adopted this system, or brought it with them, and many of the isolated hamlets on the higher slopes of the hills were probably in their origin only summer shelters.

The original settlement of Cheshire must have been, at least in part, a direct one, and not wholly an extension of local colonies from the Staffordshire side. A similarity has been noted between the Cheshire dialect in some respects and that of Norfolk, while the intervening counties differ.^[53] This may have arisen from Danish influence, and be a result of direct settlements on the coast. The maritime parts of North Wales have many old place-names to attest their settlements, and Chester appears to have been largely a Danish town during the later Saxon period. It was governed by twelve judges or lahmen, who were chosen from among the vassals of the King, the Bishop, and the Earl.^[54] As the institution of lahmen is Scandinavian, it is clear that there must have been a population of that race at Chester. Other circumstances that point to Northmen are the prevalence of family names ending in -son, corresponding to the Norse -sen, which survive in Cheshire, and the mention in Domesday Book of certain fines in the city of Chester being paid in oræ or by Danish money computation. The place-names of the Wirral district between the Mersey and the Dee show that it was occupied by the later Northmen. The discovery, however, of a runic inscription, which Stephens assigned to the seventh century,^[55] at Overchurch in the Wirral, proves that Anglians advanced into this district soon after the Battle of Chester in 613. Among the Domesday place-names that were apparently derived from those of early tribal settlers in Cheshire are Englefeld, Englelei, Inglecrost, Wareneberie, Leche, and Cocheshalle.

The Cheshire dialect, as spoken in different parts, shows certain well-marked differences in respect to vocabulary, pronunciation, and grammar.^[56] In the formation of place-names in the south of the county there was apparently little or no Danish influence. The speech in this part is broad and rough, differing in pronunciation from that of the northern part, and approaching more to that of north Staffordshire and Derbyshire. These are the counties in which the descendants of the early Anglian settlers were least disturbed by subsequent Danish inroads, and south Cheshire appears somewhat to resemble them. On the other hand, there is a clear line of difference between the local talk in south Cheshire and Shropshire, where the highly-pitched tone, the habit of raising the voice at the end of a sentence, the sharp and clearly-defined pronunciation, probably marks a Welsh element among the Shropshire people which is absent in south Cheshire.

As the settlement proceeded from east to west in the Mercian States, some of the Welsh people must have been allowed to exist among the newcomers. As far east as Buckinghamshire there was in the Anglo-Saxon period a place called Wealabroc, and in the south-west of Northamptonshire there exists still an old way called the Welsh Road. These names probably imply old frontier lines. As the advance was continued towards the present Welsh border, it is certain that here and there small areas inhabited by Welsh people in the midst of Old English settlements were left. Beyond the present border, as around Radnor, settlements of Old English or Scandian folk surrounded by Welsh people were formed. Offa’s Dyke, thrown up in the eighth century to divide the Welsh from the English, was not a strict ethnological frontier. There were some English to the west of it at the time it was made, or soon afterwards, and some Welsh to the east of it, as at Clun, Oswestry, and Cherbury, at which places early Welcheries existed, which were not governed by English customs.

It was along this border that the customs of the Old English settlers were brought into contact with the tribal customs of the Welsh. The various English customs of inheritance derived from tribal settlers have been described. In some important respects the Welsh differed from all of these. The land of the Welsh tribesmen was held by families and allotted to members of the family. On the death of the head of the family, it was first divided among all the sons. This, however, was not a final division. On the death of the last of these brothers, the land was again divided among all their sons per capita, each first-cousin taking an equal share. On the death of the last of these first-cousins, the land was again divided as before, each second-cousin taking an equal share. Land could be inherited, consequently, only by direct descent.^[57] There was no inheritance by daughters. There was no widow’s dower. No man was his brother’s heir. If a man left sons, they inherited; if he left none, the land was shared according to the tribal custom. This is of interest in reference to the custom of Dymock in west Gloucestershire, which was apparently left as an ethnological island of Welsh people. Its name is Welsh, and its custom was Welsh, for the land at Dymock passed on the death of the holder to the heirs of the body only; otherwise it reverted to the community or the lord.^[58]

The place-names Welsh Hampton, east of Ellesmere, and Welsh Bicknor and Welsh Newton, near Monmouth, tell the same story of mixed settlements. There was both an Englecheria and a Walecheria, of ancient origin, at Clun and at Cherbury in west Shropshire.^[59] There were English landholders and Welsh subtenants of ancient date in the great district of Archenfeld, west of the river Wye. It was owing to such conditions as these that the blending of race between the Old English and Old Welsh people went on. Then, as generations passed, English folk arose along the Welsh border who were partly of Welsh descent, having complexions somewhat darker than their forefathers—a physical characteristic they have transmitted to their descendants at the present day.

1. Cart. Sax., ii. 588.
2. English Diaelct Society, ‘Glossary,’ by D. Robertson, 194.
3. Stephens, G., ‘Old Northern Runic Monuments,’ ii. 1009.
4. Archæological Journal, xviii. 342.
5. Domesday Book.
6. Cleasby and Vigfusson, Icel. Dict.
7. Anglo-Saxon Chron.,
8. Testa de Nevill.
9. Cal. Inq. p.m., ii. 34, 196.
10. Elton, C. I., ‘Gavelkind,’ p. 176.
11. Ibid., p. 192.
12. Ibid., p. 229, quoting Camden and Gervase.
13. Hazlitt’s ed. of Blount’s ‘Tenures,’ p. 173.
14. Laws of Ethelred.
15. Worsaae, J. J., ‘Danes and Norwegians in England.’
16. Proceedings Soc. Antiq. Scot., xxvi. 189, 190.
17. Domesday Book, General Introduction, by Sir H. Ellis.
18. Cleasby and Vigfusson, ‘Icelandic Dict.,’ see lög-mathr.
19. Christison, D., ‘Place-Names in Scotland,’ Proceedings Soc. Antiq. Scot., xxvii. 269.
20. Elton, C. I., ‘Law of Copyholds,’ 134.
21. Rhys, J., ‘The Welsh People,’ p. 29.
22. Ellis, A. J., ‘English Dialects,’ p. 23.
23. Taylor, Isaac, ‘Greeks and Goths,’ 110.
24. Ibid., 111.
25. Cobbett, J. A., Journal Cambrian Arch. Assoc., vi. 76.
26. Ibid., Fifth Series, x. 5.
27. Ibid., vi. 76.
28. Ibid., Fourth Series, ix. 20.
29. Ibid., Fourth Series, viii. 13, 14.
30. Ibid., Fifth Series, ii. 70.
31. Rhys, J., ‘Studies in the Arthurian Legend,’ 330, 346.
32. Beddoe, J., Journal of the Anthropological Inst., xxv. 19.
33. Codex Dipl., No. 244.
34. Ibid., Nos. 426, 1359, 1365.
35. Bosworth and Toller, ‘Anglo-Saxon Dict.’
36. Ibid., and Latham, ‘Germania of Tacitus,’ 174, 179.
37. Cart. Sax., i. 306.
38. Ibid., i. 369.
39. Ibid., iii. 520.
40. Réclus, E., ‘Nouvelle Géographie Universelle,’ v. 334.
41. Codex Dipl., Index.
42. Shropshire Archæological and Nat. Hist. Soc. Trans., xi. 244.
43. Codex Dipl., Nos. 127, 143, 1251.
44. Cart. Sax., iii. 242.
45. Anglo-Saxon Chronicle.
46. Ellis, A. J., ‘English Dialects,’ 24.
47. Stephens, G., loc. cit., iii. 160.
48. Codex Dipl., No. 277.
49. Ibid., No. 1246.
50. Ibid., Nos. 637, 1366.
51. Transactions Philological Soc., 1875-1876, p. 576.
52. Bateson, M., English Hist. Review, 1901, p. 109.
53. Beddoe, J., ‘Races of Britain,’ p. 70.
54. Lappenberg, J. M., ‘History of England under the Anglo-Saxon Kings,’ ii. 354; and Domesday Book.
55. Stephens, G., loc. cit., iv. 53.
56. Darlington, T., ‘Folk-Speech of South Cheshire.’
57. Rhys, J., and Jones, D. B., ‘The Welsh People,’ 221, 222.
58. Pollock and Maitland, ‘History of English Law,’ ii. 272.
59. Plac. de quo warr., 681.