Teutonic elements as Indo-Iranic, is, according to the 181 author's expression, "the intermediate organic member" between Teutonic and Indo-Iranic. Now, if Slavo-Lithuanian is indissolubly connected, as appears from grammatical and lexical facts, in a special manner both with Teutonic and with Indo-Iranic, that is to say both with a form of European Aryan and with Asiatic Aryan; if the characteristics of the first and the second and those of Teutonic are fused together, so to speak, in Slavo-Lithuanian; if all this is true, as Schmidt thinks it is without doubt, the obvious result is that we cannot admit either a fundamental SlavoLithuano-Teutonic language (which would exclude IndoIranic), nor a fundamental Indo-Irano-Lithu-Slavonic language (to which Teutonic would be foreign), nor a fundamental European language (because Lithu-Slavonic, philologically considered, does not belong at all more to the Aryan languages of Europe than to those of Asia), nor, lastly, a fundamental Indo-Irano-Lithu-Slavo-Teutonic language (and in fact there are between the Slavonic, Lithuanian, Teutonic and the other languages of the European Aryans certain common characteristics owing to which we may not separate thus the former from the latter). So vanishes the boundary line drawn between the Aryan languages of northern Europe and those of Asia. Let us see now whether there are well defined boundaries between these and the Aryan languages of the southern part of Europe. Between the grammatical structure of Greek and that of 182 Indo-Iranic there are, especially in conjugation, relations more numerous and more important than between the Asiatic Aryan and the Italo-Keltic languages: the lexicon further Schmidt, that the oldest Teutonic documents go back to a much more remote period than the Slavonic, and especially the Lithuanian. For the rest, we know by this time that it is not advisable to assign an un

limited value to lexical comparisons.

1 On this argument see also Delbrück and Windisch Syntactische forschungen, etc. i. 102-4, and Jolly, Ein kapitel vergleichender syntax, etc., pp. 117-27. It would be

offers 99 word-forms exclusively Graeco-Indo-Iranic, and only 20 Italo-Indo-Iranic (so that the former are to the latter almost exactly as 5 to 1).' That between Greek and Italic there is a special affinity could hardly be doubted after the proofs brought forward by other philologists. Greek therefore is closely connected both with Indo-Iranic and with Italic, in other words it is the intermediate organic member between the former and the latter. There is not therefore a fundamental Graeco-Italo-Keltic language (as appears from the relations observed between Greek and Asiatic Aryan), nor a fundamental Indo-Irano-Greek language (and this results from the relations perceived between Greek and the Italo-Keltic languages): the hypothesis then of a fundamental European language is, a second time, shown to be contrary to facts. From the results of the researches prosecuted by his predecessors J. Schmidt concludes that Italic and Keltic themselves also stand as intermediate organic members between Greek and Teutono-Lithu-Slavonic. Hence, the author concludes, in place of the supposed fundamental languages we have an uninterrupted 183 series of languages by which we pass gradually from eastern Asiatic Aryan to the European Aryan of the west. As a

of little use to observe with Fick against Schmidt that certain characteristics, common exclusively to the Indo-Iranic and the Greek verb (e.g. the augment, the reduplication of several aorists) may have been lost in the other European languages: because, in the first place, such an observation would be more to the point if the question concerned words, while as it is we are talking only of forms; 2ndly, we should always be left to seek the reason why this loss should have taken place in all the European languages except Greek, if we must not have recourse to chance,

a supposition the consequences of which we have just noticed.

1 According to Fick the ratio between the former and the latter would be, on the contrary, 108 to 65, that is, less than 2 to 1. It must be added, observes Fick, that the oldest remains of Greek are greatly anterior to those of Italic, and that the latter is much less rich in words.

2 From the various formation of the infinitives an attempt has been made to raise an objection to Schmidt's doctrine: but, to say the truth, it would not be an argument very favourable to the con

whole the Indo-European languages diverged from the

primitive type in proportion to their distance from the east: two Aryan languages bordering upon one another have 184 always some feature common to them alone. There were not at first well defined boundaries between language and language in the field of the Aryan languages: two forms of the primitive Indo-European, however great was the interval between them, were united by intermediate forms without interruption. Later this complete continuity was destroyed by the mastery which, for reasons of various nature, not unfrequently one Aryan dialect gained over other cognate dialects which became lost. From what we have said it is plainly evident that Schmidt's doctrine of the relations between the Aryan languages has not, like that of Schleicher, Lottner and Fick, a genealogical, but in fact a geographical character: its symbol cannot be a tree representing the supposed successive divisions of the Indo-European stock, but rather a "wave which spreads in concentric circles ever thinner in proportion to their distance from the centre," or even "an oblique plane inclined from Sanscrit to Keltic in an uninterrupted line."

221

Such is J. Schmidt's doctrine with regard to the affinity existing between the languages of Aryan stock: such the gravest objections which have been raised against it, and the answers of the author. From the critical exposition which we have given of them, although in a very compendious form as befits this book, the result appears to us to be, that the most important among the arguments adduced by Schmidt, especially moreover the phonological argument which was the first we noticed, cannot be regarded as refuted so completely that Schmidt's theory has not the right to be considered at least as worthy of respect as the contrary doctrine. To pronounce a decisive opinion, if that will ever be possible, on such a question we must have a

1 Ebel (quoted by Schmidt) compared the Aryan languages of

Europe to a chain, the two extremities of which touch Asia.

more complete investigation of the characteristics common to two or more Indo-European languages, among which should be especially noticed the manifestly new formations which appear to be identical or similar in some of them.

We

With these reflections we conclude the second and last part of the present book. It appears from the former, as from this, that during the last decade has been continued with lively and persistent laboriousness the wonderful work of the preceding fifty years, at one time by new researches over the ground already explored, at another investigating portions not yet essayed, almost always with correct method and not unfrequently with very considerable success. must not, however, conceal the fact, that the highest problems of Aryan philology cannot yet be regarded as solved. But the conquests it has so rapidly made in the realm of truth are undoubted pledges of more splendid future triumphs. Effective instruments of victory will be found especially in the accurate examination of the Vedic dialect, the investigation of the linguistic stocks most akin to the Aryan, the study of the sciences which are most closely allied to the science of language, especially of certain portions of physiology and true psychology, the fidelity to that strict method to which modern philology owes so much, and against which it has not rebelled and cannot rebel with impunity, and that pure and foreseeing love of the truth, which dissuades us with equal force from blind faith in the results of past investigations and the inconsiderate passion for unripe innovations.

185