which would spring up by their connection with the Anglo-Saxons would be neo-Britannic rather than Teutonic. But the evidence, such as it is, for the extent of the Teutonic conquests, tells us only of the subjugation of the country, not of the occupation of the towns; and the continuance of the Roman constitution of the towns, with the prevailing characteristics of the present town population of England, may be taken as evidence that the effect of the Saxon conquests was chiefly to drive the pure British population from the country into the towns.

If it be urged that from the inferiority of the Celtic to the Teutonic race we may infer the gradual disappearance of the Briton before the Anglo-Saxon, it may be answered that this inferiority must first be proved; and that, if any inferiority be allowed, it was far less than that which marks populations such as those of New Zealand and Australia, as compared with the white colonists. On the Hindu inhabitants of India the English have made no impression, and the dif ference between the Anglo-Saxon and the Briton was not nearly so great as that which separates the Englishman from the Rajpoot or the Bengali.

The history of the Danish successes in England seems rather to militate against the Teutonic hypothesis. Those successes can scarcely be explained by the overwhelming numbers of the invaders, for the territory from which they came was much smaller than that which had been the home of the Anglo-Saxon emigrants; but the dissensions which favoured the kinsfolk of Hengst and Cerdic would now favour the countrymen of Olaf and Cnut, and the repeated treacheries of Eadric Streone can scarcely be accounted for except on the hypothesis of alliances between the Danes and the Britanno-Saxons or neo-Britannic inhabitants. But the brunt of the Danish invasions would fall on the eastern coasts, and would be scarcely felt in the western parts of the island; and thus the Teutonic population would suffer to a far greater extent than the British. If it be argued that Danish ascendency implies a considerable Danish element in the people, it may be answered that nothing in the subsequent history of the country runs counter to such a notion. The Anglo-Saxon dynasty ended practically with Ethelred, Edward the Confessor being half Norman, and Harold, the son of Godwin, half a Dane. But when all allowance has been made for the number of Anglo-Britons slaughtered by the Danes, the argument for the large preponderance of the British element in the population still seems to be unaffected.

We come, lastly, to the Norman Conquest. Judging from the number of knights mentioned in the roll of Battle Abbey, Sir James Mackintosh, following Sismondi, estimates the whole invading army at about twenty or twenty-five thousand men. But like the Roman legions, these did not belong to a single tribe or race; and the

Picards, and more especially the Britons, who followed the standard of William of Normandy, were probably far more nearly akin to the ancient Britons than to the Saxons or Danes. From the times of the Norman Conqueror to our own, the population has been affected .not so much by settlements of Flemings in Pembrokeshire, or the influx of French Huguenots, as by the immigration of Scots (whether Celtic or Lowlanders), from the days of James I., and more especially of Irishmen, who have found their way into our large towns to such an extent, that in 1851 there were 733,866 natives of Ireland living in Great Britain, as a set-off against 56,665 foreigners of all descriptions. The elements so introduced would not tend to increase the proportion of Teutonic to non-Teutonic blood in the people of England.

If in addition to these facts the areas of the Anglo-Saxon hundreds and the number of places in the several parts of the country are taken into account,' the historical evidence seems to be exhausted. The evidence of philology would probably not be adduced by any sound philologist of the present day, as a means of settling the question. The speech of a people cannot of itself determine the race to which they belong; and Mr. Pike seems rather to oppose a giant of his own creation, when he asserts that philology "will lend her aid joyously to cloak any ethnological absurdity, but shrinks away abashed from the naked truth." To Professor Max Müller, from whom he makes more than one citation, he does but scanty justice. In his first series of Lectures, Professor Müller had stated explicitly that "to the student of language English is Teutonic, and nothing but Teutonic. The physiologist may protest, and point out that in many instances the skull, or the bodily habitat of the English language, is of Celtic type; the genealogist may protest and prove that the arms of many an English family are of Norman origin: the student of language must have his own way. Historical information as to an early substratum of Celtic inhabitants in Britain as to Saxon, Danish, and Norman invasions, may be of use to him. But though every record were burned, and every skull mouldered, the English language as spoken by any ploughboy would reveal its own history, if analyzed according to the rules of comparative grammar." Here Mr. Pike joins issue with the assertion that the English

(1) Mr. Isaac Taylor estimates the proportion of the Celtic to the Teutonic population by the number of the slaves, as well as from the area of the Saxon hundreds in different countries. Sir J. Mackintosh, from the extreme inequality in the number of slaves (York and Lincoln, with two others, having none), infers that this class of men had various names in different counties, or that different sets of commissioners employed in the survey varied from each other in their language. But whatever may have been the numbers of the slaves, the civil wars of the Teutonic invaders, not less frequent or savage than those of the British inhabitants, render it difficult, if not impossible, to determine the race to which the servile class principally belonged.

language is not wholly or even in the main Teutonic; and he refers to Professor Max Müller as his authority for the statement that of the words in the language not one third portion is of Teutonic origin. It would not, however, be a sufficient answer to Mr. Pike, to say that no amount of foreign words imported into a dialect will affect the character of that dialect, because he maintains that English exhibits not merely foreign words, but foreign grammatical forms; and by way of illustration, adduces the terminations-ble, of adjec tives, and ance, ment, &c., in nouns-as distinctly of Romanic origin. Mr. Pike's argument either fails to affect the question, or goes too far. The termination of a word may be imported together with the word itself, without affecting the character of a language, unless the inflections of the foreign dialect are imported with them. The adjectival ending ble is not so much French as Latin, and the modification which it has undergone is the result of a process common to all modern Aryan languages; and the assertion that such terminations in English are instances of hybrid grammar, must be followed up by the assertion that a pure or non-hybrid grammar is not to be found. It is, of course, quite possible so to link together the resemblances between the words and grammatical forms of different languages as to reach the conclusion that all the Aryan languages are really one and the same language; and so in a certain sense they are. The very use of the word dialect involves this admission; but these dialects may be classified according to certain salient features, these being the inflexions of nouns and verbs. Nor can Mr. Pike fairly ask why in our ordinary conversation we abandoned the Teutonic brethren for the Romanic brothers, unless he is prepared also to admit that the modern High German is more closely akin to the French than is the English.

This unnecessary antagonism may, perhaps, indispose some to consider impartially the philological evidence, which, under certain circumstances, Mr. Pike does not scruple to adduce from this "dangerous ally" in illustration of English history. This evidence is found partly in geographical names, and especially in the names of rivers and hills. Far from questioning the statement that names containing the low Celtic uisge, water, point to the occupation of all Britain by a low Celtic race, "not necessarily at one and the same time, but every portion at some time or other," Professor Max Müller would probably admit that the existence of such names as Acheron, Achelous, Axius, Oxus, and Jaxartes point to a still wider extension of the same or of a kindred race. It is, of course, possible that the race which gave these names may have altogether passed away, just as in America Indian names are retained for many towns where Indian blood can no longer be traced; but in the absence of any evidence to this effect in the case of the old British people, we

seem to be justified in marking any features which the English of the present day owe apparently to their British ancestors, and not to the Teutonic invaders.

These features are to be found in the spoken, rather than in the written, language of the country; and too much stress can scarcely be laid on a distinction the force of which is fully admitted by Professor Max Müller. The popular dialect is the living speech of a nation; the literary language may be compared rather to a branch torn from its parent trunk, and its condition is one not of growth, but of phonetic decay. It follows from this that peculiarities of pronunciation may be taken as certain evidence of the present or past existence of a people from whom these peculiarities have been inherited; and if evidence be not forthcoming of the disappearance of that race, it seems fair to infer that these peculiarities survive because the main stock of the people has continued unchanged. In such cases it is better to confine our attention to certain points which are not likely to be called into question; and we may therefore ask how the presence of the letters th, as representing both the Cymric th and the Cymric dd or dh, is to be accounted for in English, while it is not found in Teutonic dialects. If it be urged that the form of the letter is found in German, the reply is that the Germans pronounce neither the one sound nor the other, and that, for them, the th in thin and then is simply the t in tear. If it be urged that the existence of the letters in Anglo-Saxon proves that the sounds were uttered by Anglo-Saxon speakers, the answer is that this assumption cannot be taken for evidence of a fact on which they are not alive to bear witness; while the existing German pronunciation raises a strong presumption against it. But on the other hand there is the present fact that both these sounds characterise the speech of Welshmen and the speech of the modern Greeks. It is but fair to ask that this fact may be allowed to carry due weight; nor may those indications be disregarded which seem to show that a Cymric dd has encroached upon the simple d of the Teutonic settlers, as in the words farthing and further, which by the rural (i.e. confessedly the more Teutonic) population, are often pronounced furder and farden. The sailors of the Yorkshire coast, Mr. Pike remarks, still speak of the sea as so many fadoms deep. On the other hand the Welsh scarcely know the sound sh, which is so much in favour with Germans; and the lower classes in this country betray a tendency to follow in such words the Welsh fashion.

In the former of these two facts we have evidence, far more conclusive than any furnished by a mere coincidence of words, for a connection of the English of the present day either with the people who still speak the Cymric dialect in Britain, or with a people to whom both Englishmen and Welshmen are akin. This connection, it may

be argued, must be one of blood, unless direct proof be adduced to the contrary; and this conclusion seems to be borne out by the fact that the language of the Britons of Cæsar's day was closely allied to the Welsh of the present time. The names of the British commanders are still familiar sounds in a Welshman's ear; Armoricas are still, as they were in Cæsar's age, regions on the sea (ar môr.). The British petorrita are explained by the Welsh pedwar, form, and rhot, a wheel.

The conclusion seems to be that the modern English retain certain modes of pronunciation which are not found among Teutonic peoples or in Teutonic dialects; that these modes are in use among Welshmen and modern Greeks; that the language of the whole island in the time of Cæsar was closely related to the extant Welsh; and that the historical evidence at our command does not require us to admit that the main body of the people is other than what it was when Volusenus first approached our shores.

But although other features might be pointed out which show the influence of Cymric forms of expression on the Romanic dialects of Gaul, and even on Teutonic idioms, it is of more importance to consider the similarity of Welsh to Greek words in instances which preclude all idea of borrowing; and in making such comparisons we must not forget that modern Greeks and Welshmen alone exhibit those peculiarities of pronunciation which the English have applied to their Teutonic dialect. We may pass by the question which relates to the identity of the names Cimmerii, Cimbri, and Cymry, with the remark that a people may not be related in blood to the race whose name they bear, and that the name Cymry in particular must be taken in a sense from which the identity of Englishmen and Welshmen must not be inferred merely because it is asserted that Englishmen are descended from a people more closely akin to the Welshman than to the German. It is possible that all may in the end be traced to a common source. Such, at least, is the belief of Professor Max Müller, who maintains that "as sure as the six Romanic dialects point to an original home of Italian shepherds on the seven hills at Rome, the Aryan languages together point to an earlier period of language, when the first ancestors of the Indians, the Persians, the Greeks, the Romans, the Slaves, the Celts, and the Germans were living together within the same enclosures, nay, under the same roof." We have, therefore, his authority for comparing the words of Celtic dialects with those of Greek dialects, and for laying stress on the fact (if it can be proved to be a fact) that many Welsh words exhibit a closer resemblance to Greek words than that which is furnished by other cognate languages. Of the instances adduced by Mr. Pike, some may have but slender warrant, others may be perhaps mistaken; but it is difficult to resist the conclusion when we compare the Welsh huddygl, soot, with the Greek ai@áλn; achlud, dark