searches on hypotaxis dispose us to believe it posterior to parataxis and derived in various ways from it. Such are the important additions and corrections which Schleicher's doctrine on the primitive and fundamental Aryan considered in its individual elements has received from the latest researches. But our exposition would be very incomplete if 184 we did not at least make mention of the general considerations on the scientific reconstruction of Aryan which may be read in Johann Schmidt's monograph Die verwantschaftsverhältnisse der indogermanischen sprachen,' with which we shall soon have to concern ourselves at greater length. There is certainly, the author observes, a series of words and forms which declare themselves to be undoubtedly Proto-Aryan. But of other lexical and grammatical elements it is no longer possible to discover the primitive form in many cases the result of this labour can only be doubtful, because, according to the learned philologist the problem has not yet been solved, "in how many languages a word-form ought to be found in use to be rightly considered as belonging to the prehistoric language of the Aryans." Moreover of the very forms which have been or can be demonstrated to be Proto-Aryan we do not know the chronology: consequently we cannot affirm that two of them are contemporaneous. "The fundamental forms may have arisen in periods altogether distinct and there is nothing as yet to assure us that the fundamental form A was still unchanged when B arose, that the forms C and D of contemporaneous origin have alsó remained during an equal time unchanged, etc. When therefore we wish to write in the primitive language a sentence," or to join several words one to another (like Schleicher in his story), "it may easily happen that such a sentence, though every element in it may be in itself well re-constructed, may

1 Weimar, 1872, pp. 28-31.

nevertheless, as a whole, be no better composed than the translation of a verse of the Gospels, the several words of which have been taken partly from the translation of Ulfilas, 165 partly from that of the so-called Tatian, partly from that of Luther, because a historical view into the primitive language is still altogether a desideratum.”

But even admitting with J. Schmidt that the fundamental language of the Aryan stock considered in its entirety is as yet nothing else than "a scientific fiction," it is nevertheless certain, even in his view, that it is a powerful help to philological investigation, nor, in our opinion, can its value be denied as a revelation of the prehistoric civilisation of the Aryans, although certainly it is considerably diminished by the lack above noticed of chronological information with regard to the development of such fundamental language. On the other hand the science of the most ancient periods could derive no advantage from the reconstruction of Proto-Aryan, or, to put it better, of Proto-Aryan roots, stems and words, unless a real existence could be attributed to the reconstructed words. It appeared that there could be no doubt about such existence, especially after the publication of the great work of Adolphe Pictet.' Against this we find J. Schmidt not disposed to admit as altogether certain the reality of Proto-Aryan except on the understanding that "the origin of the human race from some few individuals can be proved." And Sayce does not discern in the primitive Indo-European, reconstructed by science anything else than an ideal language. We should not

1 With regard to the fundamental language of the Aryans it will be well to read the reflections recently made by Bréal(La langue indo-européenne, examen critique de quelques theories relatives à la langue mère indo-européenne, in the Journal des

savants, October, 1876.)

2 Les origines indo-européennes ou les Aryas primitifs, essai de paléontologie linguistique, Paris, 1859-63. 3 Ibid., pp. 29-30.

4 The Principles of Comparative Philology, p. 126, ed. 2.

indeed know how to explain the existence of such numerous and intimate affinities as those which no one can deny between the Aryan languages, without admitting that they prove the common origin of such languages from a fundamental language, as we should not know to what cause to attribute the relations of resemblance which are seen between the Neo-Latin languages unless we traced them to the Latin which became changed into them. To 166 suppose that several nations, related indeed to each other anthropologically and geographically, but nevertheless not constituting a single society, have formed, simply by reason of the supposed affinity, each of them its own language employing the various methods which we have just mentioned, agglutination, symbolism, differentiation and adaptation, so that there resulted languages so connected with each other as the Aryan languages, appears to us much less reasonable than the hypothesis of the descent of all of them from a common mother-language spoken in prehistoric times.

§ 29. We have elsewhere described the most important results of Pictet's researches on the subject of Proto-Aryan civilisation, as it is illustrated by the primitive language which we have discussed: we have noticed also the criticisms of A. Weber and A. Kuhn on the first volume of the work of the distinguished French scholar.' Here we will observe, that a work of this kind has been taken up afresh, more in the manner of a sketch than a complete treatise, by that learned investigator Fick. And that he has arrived at conclusions not very different from those of Pictet is a fact that certainly calls for consideration, as one which ought to increase not a little our faith in the results of such researches. Fick agrees with Pictet

1 See Pezzi's Introduction, etc., pp. 210-20. 2 Die ehemalige spracheinheit der

Indogermanen Europas, Göttingen, 1873, pp. 266-85.

N

also in discovering indications of monotheistic thought in the midst of the naturalism, rich in deities, of the primitive Aryan people.

But in the last decade there has been a notable disagreement between the illustrious author of the treatise on Les origines indo-européennes and certain philologists in the quest after the fatherland of the Aryans. While Pictet 167 with the majority of the scholars of our time taught that the Aryan people, as yet undivided, occupied in prehistoric times an Asiatic district, of which the Bactrian district may be regarded as the centre, Benfey and Geiger3 with some other scientific men sought in Europe the home of the primitive Indo-Europeans. Moreover that most learned Iranic scholar, Spiegel, does not deny credit to the possibility of the European origin of the Aryan people.* The arguments adduced to support this new hypothesis are derived partly from anthropology (and with these we are not concerned), partly from the science of language: to these last belongs that which Benfey discovers in the lack of names common to all the families of Aryan languages for the great wild animals of Asia, for example the lion, which, as is well known, existed in Greece even in historic times.5

1 Höfer, Die heimat des indogermanischen urvolkes (in the Zeitschr. f. vgl. sprachforsch, xx. 37984).

2 See the preface to Fick's Wörterbuch der indo - germanischen grundsprache (Göttingen, 1868) and the Geschichte der sprachwissenschaft, etc., pp. 599-600.

3 Zur entwickelungsgeschichte der menschheit, Stuttgart, 1871, pp. 113-50 (quoted by Höfer in the monograph noticed).

4 Eránische alterthumskunde, i. Leipzig, 1871, pp. 426 sqq. (quoted by Höfer, ibid.).

5 Pauli (Die benennung des löwen bei den Indogermanen, ein beitrag zur lösung der streitfrage über die heimath des indogermanischen urvolkes, Münden, 1873) subjects to examination this argument, and rejects it as devoid of force, arriving at other conclusions from the study of the names by which we see the lion denoted in the IndoEuropean languages. His work has been made the subject of critical reflections by Wolzogen, according to whom the various denominations of the lion in the neighbourhood of the Aryan peoples cannot furnish any

But the Proto-Aryan names of such animals might well have been lost, especially in countries so distant and so different from those in which the fatherland of the Aryans is generally imagined to be. To the merely negative and therefore very weak argument of Benfey may be opposed the certain and important fact that the Asiatic forms of the Aryan languages are generally much closer to the primitive type than the European forms: this fact, which finds its counterpart also in comparative mythology,' is in 188 our opinion a very weighty indication that the Aryans. migrated from Asia into Europe, not from Europe into Asia. We are well aware that, in order to combat this formidable argument, it has been observed that the archaic form of the Vedic is not a sufficient proof in favour of the Asiatic origin, because it must be regarded as a sacerdotal dialect. But the truth is that we could not reconcile this statement with the results of the latest linguistic studies of the Vedas: these studies have put in a clear light how in those very ancient records of the thought and the word of the Indo-Europeans there are still seen not unfrequently traces of the natural processes by which the languages of Aryan stock became perfected. Nor is it of any use to object that we might perhaps discover not less numerous, not less important vestiges of antiquity in Aryan languages of Europe, for example in Greek, if records of these had reached us belonging to epochs not much later than the period to which the Vedas are referred. And in fact, every language exhibits, especially in its phonetic and its formal portion, certain characteristics so deeply marked that time cannot either cancel or change them as much as it would be necessary to suppose in order to admit the hypothesis

certain index to the country inhabited by the Indo-Europeans before their division (Zeitschr. f. völkerpsychologie, etc., viii. 206-15).

1 Wolzogen, Der ursitz der Indogermanen (Zeitschr. f. völkerpsychologie, etc., viii. 1-14).