In the cities, indeed, it would seem that the Britons remained in great numbers. The Welsh bards complain that the urban race of Romanised natives known as Loegrians, "became as Saxons." Mr. Kemble has shown that the English did not by any means always massacre the inhabitants of the cities. Mr. Freeman observes, "It is probable that within the [English] frontier there still were Roman towns tributary to the conquerors rather than occupied by them;" and Canon Stubbs himself remarks, that "in some of the cities there were probably elements of continuous life: London, the mart of the merchants, York, the capital of the north, and some others, have a continuous political existence." "Wherever the cities were spared," he adds, "a portion, at least, of the city population must have continued also. In the country, too, especially towards the west and the debateable border, great numbers of Britons may have survived in a servile or half-servile condition." But we must remember that in only two cases, Anderida and Chester, do we actually hear of massacres; in all the other towns, Bæda and the Chronicle tell us nothing about them. It is a significant fact that Sussex, the one kingdom in which we hear of a complete annihilation, is the very one where the Teutonic type of physique still remains the purest. But there are nowhere any traces of English clan nomenclature in any of the cities. They all retain their Celtic or Roman names. At Cambridge itself, in the heart of the true English country, the charter of the thegn's guild, a late document, mentions a special distinction of penalties for killing a Welshman, "if the slain be a ceorl, 2 ores, if he be a Welshman, one ore." "The large Romanised towns," says Professor Rolleston, "no doubt made terms with the Saxons, who abhorred city life, and would probably be content to leave the unwarlike burghers in a condition of heavily-taxed submissiveness."
Thus, even in the east it is admitted that a Celtic element probably entered into the population in three ways,—by sparing the women, by making rural slaves of the men, and by preserving some, at least, of the inhabitants of cities. The skulls of these Anglicised Welshmen are found in ancient interments; their descendants are still to be recognised by their physical type in modern England. "It is quite possible," says Mr. Freeman, "that even at the end of the sixth century there may have been within the English frontier inaccessible points where detached bodies of Welshmen still retained a precarious independence." Sir F. Palgrave has collected passages tending to show that parties of independent Welshmen held out in the Fens till a very late period; and this conclusion is admitted by Mr. Freeman to be probably correct. But more important is the general survival of scattered Britons within the English communities themselves. Traces of this we find even in Anglo-Saxon documents. The signatures to very early charters,[3] collected by Thorpe and Kemble, supply us with names some of which are assuredly not Teutonic, while others are demonstrably Celtic; and these names are borne by people occupying high positions at the court of English kings. Names of this class occur even in Kent itself; while others are borne by members of the royal family of Wessex. The local dialect of the West Riding of Yorkshire still contains many Celtic words; and the shepherds of Northumberland and the Lothians still reckon their sheep by what is known as "the rhyming score," which is really a corrupt form of the Welsh numerals from one to twenty. The laws of Northumbria mention the Welshmen who pay rent to the king. Indeed, it is clear that even in the east itself the English were from the first a body of rural colonists and landowners, holding in subjection a class of native serfs, with whom they did not intermingle, but who gradually became Anglicised, and finally coalesced with their former masters, under the stress of the Danish and Norman supremacies.
In the west, however, the English occupation took even less the form of a regular colonisation. The laws of Ine, a West Saxon king, show us that in his territories, bordering on yet unconquered British lands, the Welshman often occupied the position of a rent-paying inferior, as well as that of a slave. The so-called Nennius tells us that Elmet in Yorkshire, long an intrusive Welsh principality, was not subdued by the English till the reign of Eadwine of Northumbria; when, we learn, the Northumbrian prince "seized Elmet, and expelled Cerdic its king:" but nothing is said as to any extermination of its people. As Bæda incidentally mentions this Cerdic, "king of the Britons," Nennius may probably be trusted upon the point. As late as the beginning of the tenth century, King Ælfred in his will describes the people of Devon, Dorset, Somerset, and Wilts, as "Welsh kin." The physical appearance of the peasantry in the Severn valley, and especially in Shropshire, Worcestershire, Gloucestershire, and Herefordshire, indicates that the western parts of Mercia were equally Celtic in blood. The dialect of Lancashire contains a large Celtic infusion. Similarly, the English clan-villages decrease gradually in numbers as we move westward, till they almost disappear beyond the central dividing ridge. We learn from Domesday Book that at the date of the Norman conquest the number of serfs was greater from east to west, and largest on the Welsh border. Mr. Isaac Taylor points out that a similar argument may be derived from the area of the hundreds in various counties. The hundred was originally a body of one hundred English families (more or less), bound together by mutual pledge, and answerable for one another's conduct. In Sussex, the average number of square miles in each hundred is only twenty-three; in Kent, twenty-four; in Surrey, fifty-eight; and in Herts, seventy-nine: but in Gloucester it is ninety-seven; in Derby, one hundred and sixty-two; in Warwick, one hundred and seventy-nine; and in Lancashire, three hundred and two. These facts imply that the English population clustered thickest in the old settled east, but grew thinner and thinner towards the Welsh and Cumbrian border. Altogether, the historical evidence regarding the western slopes of England bears out Professor Huxley's dictum as to the thoroughly Celtic character of their population.
On the other hand, it is impossible to deny that Mr. Freeman and Canon Stubbs have proved their point as to the thorough Teutonisation of Southern Britain by the English invaders. Though it may be true that much Welsh blood survived in England, especially amongst the servile class, yet it is none the less true that the nation which rose upon the ruins of Roman Britain was, in form and organisation, almost purely English. The language spoken by the whole country was the same which had been spoken in Sleswick. Only a few words of Welsh origin relating to agriculture, household service, and smithcraft, were introduced by the serfs into the tongue of their masters. The dialects of the Yorkshire moors, of the Lake District, and of Dorset or Devon, spoken only by wild herdsmen in the least cultivated tracts, retained a few more evident traces of the Welsh vocabulary: but in York, in London, in Winchester, and in all the large towns, the pure Anglo-Saxon of the old England by the shores of the Baltic was alone spoken. The Celtic serfs and their descendants quickly assumed English names, talked English to one another, and soon forgot, in a few generations, that they had not always been Englishmen in blood and tongue. The whole organisation of the state, the whole social life of the people, was entirely Teutonic. "The historical civilisation," as Canon Stubbs admirably puts it, "is English and not Celtic." Though there may have been much Welsh blood left, it ran in the veins of serfs and rent-paying churls, who were of no political or social importance. These two aspects of the case should be kept carefully distinct. Had they always been separated, much of the discussion which has arisen on the subject would doubtless have been avoided; for the strongest advocates of the Teutonic theory are generally ready to allow that Celtic women, children, and slaves may have been largely spared: while the Celtic enthusiasts have thought incumbent upon them to derive English words from Welsh roots, and to trace the origin of English social institutions to Celtic models. The facts seem to indicate that while the modern English nation is largely Welsh in blood, it is wholly Teutonic in form and language. Each of us probably traces back his descent to mixed Celtic and Germanic ancestry: but while the Celts have contributed the material alone, the Teutons have contributed both the material and the form.
[1] The word in the original is exterminatis, but of course exterminare then bore its etymological sense of expatriation or expulsion, if not merely of confiscation, while it certainly did not imply the idea of slaughter, connoted by the modern word.
[2] In this and a few other cases, modern authorities are quoted merely to show that the essential facts of a large Welsh survival are really admitted even by those who most strongly argue in favour of the general Teutonic origin of Englishmen.
[3] Kemble "On Anglo-Saxon Names." Proc. Arch. Inst., 1845.