exactly on the surface (inasmuch as they differ from those given by the text of Beda), lie but a little below it: doctrines which most independent investigators, more or less, agree with-some having anticipated, some adopted them. On the other hand, some prefer the older views. In saying this, however, we may add that their number decreases yearly for it may safely be asserted that the belief in Hengist and Horsa and their congeners is dying out. However, even if it be not, there is nothing in anything which has preceded which ought to be mistaken for a paradox. The doctrine that a great many changes are nominal is (perhaps) opposed to the opinions of several influential writers in Germany, who believe in real displacements much more than does the present investigator.

§ 175. So much for what has gone before. What follows is in a different category. It is not wholly speculative if it were so, it should find no place in an elementary work like the present. Nevertheless, it is more speculative than critical-though not without critical elements. In some places it is more ethnological than philological: but ethnology and philology are allied. Upon the whole, however, it is meant to be suggestive of further research rather than authoritative for what is already admitted. Few things, in rerum naturá, are, at one and the same time, very true and very original; inasmuch as, when once many men are thinking on the same subject, and are moving in the right direction, there is a great deal of unconscious agreement. At the same time, the contents of the present chapter are those which (to say the least of them) their expositor holds in opposition to many with whom he is unwilling to differ.

§ 176. Are any broad and trenchant lines of demarcation between the several divisions of the German group of languages tenable ?-If they be not, the division between the

Teutonic and the Norse branches is of small import; and Southern Danish may be much more akin to Northern Frisian, or Northern Angle, than the ordinary questions about Danish, as opposed to Angle, elements suggest. This, however, is a special detail. It is here submitted that, if German philology is to be advanced, and if we are to emancipate ourselves from the influence of names, our groups must be made according to type rather than according to definition. This means that extreme forms, along with those that approach them, can only be separated by the latter method. When we approach the confines, one class graduates into another. Low-German is easily separated from High-German, as long as we take one or two tests. But many of them are arbitrary. Now, unless we are satisfied with these, the lines between the Frank and the Saxon, like those between the Saxon and the Frisian, are indistinct. Upon the Frisian and the Norse more will be said in the sequel. At present it is enough to state that classification in the way of definition is, in many cases (I do not say in all, or even in the majority), only practicable when we take either extreme forms or single characteristics. As a single example-the Dutch of Holland and the English of England are descendants from different members of the German family; yet the Carolinian Psalms have been claimed by both Dutchmen and Saxons.

§ 177. Original magnitude of the German areaThe Goths other than German.-Upon two points I have found reason to differ with, at any rate, the majority of inquirers; perhaps, with all of them. I find no reason to believe that the original occupancy of the Germans was at any earlier date different from what it was at the beginning of the true historical period, i. e. the time of the Carlovingians; when the Franks were pressing upon the Saxons and the Frisians, and when an

incipient Christianity was, for the first time, supplying us with trustworthy first-hand observations. The Germans as they were found by Adam of Bremen and his contemporaries I believe to have been the Germans of the time of Tacitus. But, at the time in question, nothing was German from the Elbe eastward. On the contrary, everything was Slavonic. The details of this doctrine I have exhibited elsewhere.

Again-though the term Goth is held to be nearly synonymous with German, I find no evidence of any German whatever having been called, by himself or by any one else, a Goth until he had settled on the land of the Geta or Gothi. If so, he was a Goth in the way that an Englishman is a Briton, a Spaniard a Mexican, or a Portuguese a Brazilian, i. e. not at all. If so, the whole early history, not only of the Goths, but of every nation whose name has been identified with Goth (the Jutes, Gothini, &c.), must be transferred elsewhere. The details of this view have also been given elsewhere. Elsewhere, too, have been given reasons for believing that the real Goths were Lithuanians upon Fin soil, i. e. Lithuanians in the way that a German is a Briton. If so, the term implies, in most cases, three strata of population; (1) Fin, or Ugrian, (2) Lithuanic, (3) German. If so, the Germans of Scandinavia were not the second occupants of Scandinavia, but the third; a Lithuanic and, in some cases, a Slavonic immigration having intervened between them and the earlier Fins. I believe that this doctrine is not without its adherents in Norway, Denmark, and Swedenfew though they be. This, however, is by no means incompatible with a disbelief in the so-called Fin hypothesis; by which we are taught that, before the westward movement of the so-called Indo-Europeans, everything in Europe, and a great deal elsewhere, was Fin, or Ugrian. It is one thing to believe, as a matter of comparatively

recent history, that the Fins of Estonia, Liefland, and Curland extended beyond the Oder, and possibly beyond the Elbe and Eyder; another to maintain that they were the aborigines of Bohemia, Gaul, or Spain.

§ 178. Early Sarmatians, i. e. Slavonians or (and) Lithuanians.—Something may be said in favour of the Picts having been Sarmatians. The whole subject, however, of the Pict nationality is difficult. There is some objection to every hypothesis. Independently, however, of the Picts, I think that Slavono-Lithuanic descents upon the British islands, during the time when the Sarmatian seaboard along the Baltic extended to the Trave, are, by no means, unlikely indeed, I hope, in a fitter treatise than the present, to give reasons for believing such to have been the case.

§ 179. Can the Angles have been less German than their language makes them?-Yes. In whatever way we interpret the fact upon which so much stress has been laid, viz. the extension of the Slavonic area to and beyond the Elbe, we must see the great probability of German and Slavonic intermixture. It is probable, if the Slaves pressed forward into Germania. It is probable, if the Germans encroached upon Slavonia. I would not say that the word Angraria may not be, like Ingria, Wagria, Ukr-in, Ucker-mark, and Ukrain, an actual Slavonic gloss, march or boundary. If so, all Ostphalia may, originally, have been Slave. Indeed, when we remember that Luneburg was Slavonic, it is nearly so from the ordinary point of view.

The Lombards' relations to the Angles have already been noticed. For believing the Lombards, or Langobards, to have been Germanized Slaves, there is, inter alia, the special statement of Paulus Diaconus that they were originally called Winili =Wends. Again-the nation with which the few notices of the Angles of Germany most

particularly connect them is that of the decidedly Slavonic Werini, or Varini. If this be the case, our belief in the agreement between the language of a population and its blood must be very different from that of the great upholders of what may be called the pre-eminence of the Anglo-Saxon race. To say nothing of the amount of British blood which they may have taken up in Britain, our ancestors were, even in Germany, German in language, though Ugrian and Sarmatian in blood. The same applies to the Danes and Norwegians. This, however, is a point of ethnology rather than philology.

§ 180. Are the differential characters of the Scandinavian languages either recent or exotic?-What is the value of the chief characteristics of the Norse languages? What are they?

=

1st. The neuter in -t, as skön pulcher, pulchra, and skön-t = pulchrum: in which the Norse is distinguished from all the German forms of speech, save and except the Moso-Gothic, where blind-s=cæcus, and blind-ata=cæcum. That this, however, is the form out of which the Modern German -es has been developed is shown by the O. H. G. forms blindaz blindats (there or thereabouts). The pronouns retain the t throughout; but the adjectives only in Norse and Moso-Gothic. To account for this we must suppose that the Norse became distinguished from the other German tongues before the change set in. This, however, is improbable in the eyes of those who make Moso-Gothic High-German: which Mr. Kemble (for one) does not. He, especially, calls it a Low-German dialect. It might have sufficed if he had said that it was as much Low as High; as much akin to the dialects of Hanover and Westphalia as to those of Bavaria and Suabia; and I doubt whether he meant to say more. any rate, the Moso-Gothic is not, in the eye of every competent authority, High-German.

At