the human race to which we belong ever passed through the stages here dubiously indicated of monosyllabism and agglutination-stages which I hope to make clearer in another Lecture-before it attained the inflectional character.* This original speech has of course been dead for ages, and even Sanskrit, its oldest and purest representative, is dead, and Keltic, its next oldcst representative, is dying; but from comparison of all its representatives-though it perished long before history began, though no vestige of it on rock, or pyramid, or gem, or coin, or poem preserved by immemorial tradition, now remains-its forms can be conjecturally restored. To restore them in this manner was the object of the celebrated Compendium of Professor August Schleicher. But even this language is only reproducible in its perfect and fullgrown condition. How it grew we know not. No man saw the Tree planted—no mortal hand watered the bursting of the grove: no register was kept of the gradual widening of its girth, or the growing circumference of its shade, till the unexpected bole stands forth in all its magnitude, carrying aloft in its foliage the poetry, the history, and the philosophy of heroic peoples.'

But although the labours of

* It is remarkable that Dante, rising superior to the prejudices of his age, attached no faith to the popular misconception that Hebrew was the primitive language. 'La lingua ch' io parlai fu tutta spenta' are the words which he puts into the mouth of Adam. (Paradiso, canto xxvi.)

recent scholars have recovered for us, with tolerable certainty, many words of this ancient tongue, yet all that we know of it historically is that, in very early times, a great split in it must have taken place in consequence of the westward divergence of two great divisions of the hitherto united race which spoke it. To the division which occupied the northern, central, and eastern parts of Europe, we may give the name of Letto-Slavo-Teutonic, derived from the nationalities into which it was afterwards differentiated. To the division which occupied the southern and western peninsulas of Europe we may give the name of Græco-Italic. Earlier, perhaps, than either of these great divisions, the ancestors of the once wide-spread and mighty race of the Kelts had wandered into Europe. The area over which Keltic names are found diffused shows the original extent of their dominion; but they were gradually dispossessed of its central regions by the advancing Teutons, before whom they have constantly retired to the westward, and before whom their remote Irish descendants are still migrating beyond the Atlantic. For a long period after the first beginning of this westward Exodus, the Aryans proper, i.e. the ancestors of the Persians and Hindoos, were still lingering in or near their old Iranian home,* confined there partly perhaps by their love

*These general results are represented in the table of Aryan languages which accompanies this Lecture. The main conclusion

and reverence for it, and partly by the girdle of deep rivers and mighty ranges of snowy hills which barred its southern and eastern boundaries.

Now as all these events took place in the prehis toric periods of this race, you may naturally ask the grounds on which we rest such inferences, or why we represent them as being in any degree probable.

may be tabulated as follows. Similar linguistic trees may be found in Schleicher, Die Deutsche Sprache, 81, and Compend. d. vergl. Gramm. 7; and a table like the larger one is furnished in Dr.

Boltz's excellent little book, Die Sprache und ihr Leben. I differ, however, from these authorities in several important particulars, and especially in the position assigned to the Keltic languages.

The question is a very natural one, and I answer at once that the proofs are almost entirely linguistic in their character. They rest generally on the fact that certain roots-such, for instance, as those which express the numerals, the pronouns, the domestic animals, the near degrees of relationship, and other early and necessary conceptions-are (as we have already observed) common to every branch of this great family; whereas other roots are common only to the western or the eastern members of it, showing most distinctly that there was a certain heritage of roots and linguistic ideas common to the entire undivided race, while others could only have been developed separately, as occasion for them arose, long after the family had been split asunder.

If, for instance, we examine the names of plants and trees in these Aryan languages, we find the generic name for tree, branch, stem, &c., common to them all; but when we come to the specific nomenclature, we find words running through all the European family, which are totally distinct from those of Hindostan. In Hindostan the Aryans encountered a tropical vegetation, entirely unlike the temperate one to which they had been accustomed on the Iranian plateau. There was but one tree which they recognised; it was the tree which for so many centuries of English education was regarded as the necessary tree of knowledge-the tree which, if I may be pardoned the allusion, has so often

'blushed with patrician blood'—I mean the awful and venerable birch. Bhurrja is the Sanskrit name for birch;' and as it was the only tree which the Aryans, coming as conquerors from the North, were able to recognise, it is also the only tree whose name is common to Sanskrit and the languages of Europe.* A similar argument may be derived from the root lin in linen.' We indeed have adopted from Anglo-Saxon the word flax, which is derived from the same root as the Greek Tλékw, I weave; but in nearly all the European languages we find for flax such words as the Greek λívov, the Latin linum, Gothic lein, the Irish lin, the Welsh llin, and the Russian lenu; yet in the languages of India, early as the cultivation of flax was known, we find for its name such wholly different roots as atasi and ûma; showing clearly that the Western Aryans must have known and used this plant while they were a yet undivided body, yet after the great split which separated them from the Aryans of the east. And this indeed is but one out of many concurrent indications which all tend to prove the remarkable and interesting conclusion that the Eastern Aryans continued to be mainly a pastoral race, long after agriculture had been greatly developed among their brethren of the west. We should arrive at

* Klaproth's Nouv. Journ. asiat. v. 112. The root is connected with the English bark, Scandinavian börk, Gothic brikan, Greek phy-vvu, Latin frango, &c. See Pictet, Orig. ind.-eur. p. 218.

The vocabulary of agriculture is not common to the whole

G