been able to reduce to order, and combine under one law, the several insulated facts incidentally mentioned by philologists. I shall here add a few of them, which I have collected from various authors; but as they would present only a mass of confusion without some clue to guide us, I must premise that the Medo-European languages consist of the Sclavonian, Lithuanian, Latin, Gothic, Low German, and Erse; and that High German, Greek, and Welsh form the Perso-European class.

Frederick Schlegel states, that Low German is highly deserving of the attention of philologists, as it retains more of Sanskrit forms than High German": to the same purpose is the observation of Bopp, p. 35, that the oldest forms of German (the Gothic) are more similar to Sanskrit than to Persian.

"Let me observe," says Mr. Halhed in his Bengal Grammar," that as the Latin is an earlier dialect than the Greek as we now have it, so it bears much more resemblance to the Sanskrit both in words, inflexions, and terminations." Sir W. Jones too, speaking of a work which he rendered into English, informs us that he began with translating it verbally into Latin, "which," he adds, “bears so great a resemblance to the Sanskrit, that it is more convenient than any other language for a scrupulous interlineary version."

"It is a curious phenomenon," says Arndt, on the European languages, p. 88, "that the High German, which has ever been cultivated after Roman models, should approach in its structure much nearer to the Greek; whilst the Russian, which in earlier times was formed upon the

10 Uber die Sprache und Weisheit der Indier, p. 8.

Greek, bears a striking similarity to Latin: the explanation of this phenomenon must be sought in remote antiquity." In another place, p. 106, he says, that "the Latin and Sclavonian words which occur in German, belong in much greater proportion to the Low German than to the High German dialects."

"To show the German origin of the Latin language, we must not confine ourselves," says Jakel, "to the present High German, but must include the sister dialects of Sweden, Denmark, Iceland, Holland and England; but more especially the mother tongues, the Gothic, Frankish, and Anglo-Saxon: these differ in many respects from our modern High German, but at the same time they approach infinitely nearer to the Latin: indeed, it may be laid down as a general position, that the Low German has a much closer affinity to Latin than the High German ""

By referring to the table of languages given above, even the reader to whom this subject is new, may perceive an unity of principle in these various remarks, and see the foundation on which they all rest. The facts themselves not only receive light from the division of the European idioms into two classes, but also afford evidence to the reality of such a distinction.

Herodotus plainly intimates that the various tribes from the head of the Adriatic, and north of the Danube, were of Median origin. He says, that nothing certain is known concerning the people in the north of Thrace. The country beyond the Danube is a wild and undefined space; its only inhabitants, as far as I have been able to learn, are the Sigynnæ, a people who in dress resemble the Medes; their

11 Der Germanische Ursprung der Lateinischen Sprache und des Romischen Volkes, p. 16.

horses are small, but when yoked to a car, they are remarkable for their speed, for which reason cars are very common among them. The boundaries of this people extend almost to the Heneti on the Adriatic. They call themselves a colony of the Medes: how Medes reached this country, I am not able to say; though, in a long course of time, it is quite possible. (v. 9.)

It appears from Strabo, lib. xi., that the very oldest authors, long before the age of Herodotus, drew the same line of demarcation as this historian; for it was their practice to comprehend all tribes to the north of the Adriatic, Danube, and Euxine, under the general name of Hyperboreans, Sauromatæ, and Arimaspi 12. Now the Sauromatæ are generally allowed to have been an ancient branch of the Sclavonian family; and Diodorus Siculus (ii. 89.) expressly assigns them a Median origin: the title of Hyperboreans seems to have been a more general one, but I shall afterwards show that there are very particular reasons for comprehending under it the Lithuanians and Old Prussians, whose dialects are closely allied to the Sclavonian and Zend. As to the Arimaspi, they were only a particular tribe of the Hyperboreans: Αριμασποι, εθνος Υπερβορέων, Steph. Byzant.

If, therefore, a line be drawn from the head of the Adriatic across to the Danube, and along the course of that river to the Black Sea, we perceive that all the tribes beyond that line were considered by the ancients as genuine descendants of the Medes: the tribes that fall within that line were exposed to an admixture with Persian races, in times perhaps beyond the reach of history,

12 See Ritter's Vorhalle, p. 464.

but we discover marks of the event in the languages which arose from it.

This account may suffice to render intelligible, and to establish in a general way, my arrangement of the European idioms under two great classes of distinct but kindred origin: additional arguments, in support of this division, will be brought forward during the progress of the work.

For facility of reference, I shall repeat here the list of the two sets of languages :—

Medo-European --Sclavonian, Lithuanian, Latin, Low German, and Erse.

Perso-European :-Greek, High German, and Welsh.

CHAPTER III.

TABLE OF LANGUAGES:

PREFIXES:

GRIMM'S LAW:

RELATIVE ANTIQUITY OF EUROPEAN LANGUAGES.

In this chapter I propose to present the reader with an outline of the countries over which he will have to travel, and which occupy the extensive space lying between the western extremities of Europe and the river Indus. Some marks also will be pointed out, which will enable him, at any point of his wide course, to know something of the people among whom he may have arrived. To the Geologist, the materials of the road on which he is passing, or the outline of the hills which may rise in the distance, will often afford a general idea of the nature of the district in which he then is, and show the relative antiquity of its formations compared with those of the countries which he has left behind; as the appearance of a cross or a crescent on the public edifices would make known the religion of the people. In like manner, some means of forming a judgment concerning the people with whom he is at the moment conversant, are supplied to the Philologist in the particular

D