n, due to the disappearance, as Budenz holds, of t, the plural sign. In Aryan the suffix of the first person plural is ma-s (ma-si), compounded of ma, I, and the plural suffix. In Finnic the suffix of the second person plural is t-te (as in tule-t-te, ye come), compounded of the plural suffix t as before, and to thou. In Aryan the suffix was originally ta-si from ta, and the plural suffix. We see the Finnic plural suffix t which was probably the archaic form of the Aryan plural suffix s. It will be noted that the order of the signs of the plural and the pronoun is different in Aryan and Finnic. They were independently formed, after the separation of the races, but the materials out of which they were formed were identical.

The case

It is the same with the declension of the noun. signs in Finnic arose out of suffixed prepositions as in Aryan languages. We have the ablative in -t, the genitive in -n, and the accusative in -m. Thus in Tscheremis we have the accusative vida-m, water, from the stem vid-a, water, and in Sanskrit pati-m, master, from the stem pati.

In Aryan, as in Finnic, there are internal vowel changes in the stems as in Finnic, but these, probably, may date from a later period.

I feel bound to give full prominence to the two strongest arguments against the primitive identity of the Finnic and Aryan tongues, arguments that to many will, perhaps, seem conclusive against my contention.

These arguments are morphological, and seem to go down to the very foundations of grammar.

They are, first, that the Finnic languages, like the rest of the Turanian class, possess no gender; and, secondly, that the sign of the plural is inserted between the stem and the pronominal or postpositional suffixes, instead of after them, as in Aryan languages. This is also the case with Georgian, where bi or ni, the plural sign, is inserted between the root and the case endings.

I am inclined to believe that these two peculiarities of Finnic grammar, instead of being fatal to my proposition, afford a very curious confirmation of some speculations of Prof. Sayce, as to the earliest form of Aryan speech, and, therefore, if his speculations be sound, they afford a remarkable confirmation of my theory. Not only has gender been lost in two Aryan languages, English and Persian, but Prof. Sayce considers that gender did not exist in the primitive Aryan speech, in which case its absence from Finnic is only an additional proof that Aryan was derived from Finnic. In his article on grammar in the "Encyclopædia Britannica," Prof. Sayce observes that "Gender is the product partly of analogy, and partly of phonetic decay."

"There are many indications," he continues, "that the parent Aryan, at an early stage of its existence, had no signs of gender at all." "The terminations of father and mother, pater and mater for example, are exactly the same." Feminines like humus and ódós, or masculines like advena and Toλiτηs, show there was a time when these stems indicated no particular gender, but owed their subsequent adaption, the one to mark the masculine, and the other to mark the feminine, to the influence of analogy." If this reasoning is correct, and I confess I do not see any flaw, we should expect to find the parent Aryan genderless like the Finnic.

If Prof. Sayce is right, the very fact that Finnic is without gender, is one reason the more why we may look to Finnic as the parent of Aryan speech.

The same reasoning holds as to the difference in the formation of the plural. Prof. Sayce considers that in the primitive Aryan speech there was no plural, but only the singular and the dual. Now, though the plural is differently formed from the same elements in Aryan and Finnic, the dual is formed in precisely the same way. Hence I take the different formation of the Aryan and Finnic plural to be a sign of primitive unity. Prof. Sayce says1: "We might think the roots of the plural go down to the beginnings of language, but it is not so." He thinks this is proved by the existence of the dual, which would have been needless if the plural had been in existence, as we see by the fact that the existence of the plural has caused the dual to be dropped. "The dual," he says, "was older than the plural, and after the development of the latter, survived only as a useless encumbrance, which most of the Aryan languages contrived to get rid of." The same was the case with the Finnic languages, which originally had a dual, as proved by its existence in Ostiak, Lapp, and Samoyed, but the more cultured languages have got rid of it. Now, the curious point is that, though the Aryan and Finnic languages differ fundamentally in the formation of the plural, they agree precisely as to the formation of the dual.

2

The Aryan dual is believed to have been formed by two suffixed pronouns, as-ma (= I + he) being equivalent to "we two," and tas-ma (= thou + he) = ye two. In like manner Pott considered the Samoyed dual was originally equivalent to I + he, and the same holds probably of Ostiak and Lapp. The dual suffix in Finnic follows the case ending and pronominal suffix as in the Aryan languages.

1 Sayce, "Principles," page 258.

2 "Encyclopædia Britannica," article Grammar.

In the Finnic languages the dual is formed like the Aryan dual. The case ending comes first, and the sign for the dual after it.

But the Aryan and Finnic languages must have separated when they were in the stage which Prof. Sayce assigns to the oldest Aryan speech, that is, when they possessed only a singular and a dual.

In both the plural was a subsequent formation, and was formed in Aryan on the model of the dual, either by the addition of a plural suffix, or as some grammarians hold, by an intensification of the dual, while in Finnic it was formed by a plural suffix t inserted before the pronominal suffix. The singular and plural were regarded as independent words, and the suffixes were tacked on, just as in English we tack on the sign of the genitive in such words as man and men, e.g., "the man's boots," and "the 'men's boots," a formation which corresponds exactly to the formation in the Finnic languages.

I maintain, therefore, that the two chief fundamental differences between Aryan and Finnic grammar, namely, gender and the plural, instead of being proofs of primitive diversity, are, in the light of the most recent speculations, convincing proofs of primitive unity, and also that Finnic grammar is able to cast unexpected light on the primitive grammar of the holo-ethnic Aryan race.

The grammar of such a Turanian language as the Turkish seems to have no points of agreement with the grammar of the more advanced Aryan languages, such as Persian or English, but the grammar of the more advanced Finnic languages, such as Suomi or Esthonian is not far removed from that of the more archaic Aryan languages such as Sanskrit or Lithuanian, and hence the Finnic forms the link between Aryan and Turanian speech. We find a gradual progression from Buriat through Yakut and Uigur to the Tschuwash, which are all languages of the Turko-Tatar class. The Tschuwash is not very far removed from the Ugric branch of the Finnic tongues, so that through Magyar, Ostiak, Wogul, and Mordwin, we reach the Suomi and Esthonian, through which we get the transition to Lithuanian and Sanskrit, which are inseparable from the Keltic, Latin, Greek, Slavonic, and Teutonic tongues. Just as the Finnic is a development of the Turkic, so the Aryan is a development of the Finnic Ursprache.

Twenty years ago when Weske pointed out the grammatical analogies between Finnic and Aryan, he refrained from affirming that they point to a single primitive Ursprache, because at that time the primitive verbal roots of the Finnic language had not been determined. This, however, has now been done by

Budenz, Donner, and Vambery, and we can carry Weske's argument a step further, and show not only that the grammar is fundamentally identical, but the primitive roots, the Stoff out of which the voca bularies have been manufactured, is the same.

To demonstrate this proposition would require a volume. I will take one leaf only out of the book, as a sample of the rest. It will be better to examine thoroughly a small portion of the domain, than to scamper over the whole ground. Lest I should unconsciously pick my evidence, I will take a few roots in consecutive alphabetical order. Prof. Skeat, in his "Etymological Dictionary," has given a list of 461 primitive Aryan roots, mainly from Fick. Of these I have taken the 18 triliteral roots in k, Nos. 41–58, and have compared them with the Finnic k-stems in Donner's Vergleichendes Wörterbuch der Finnische Sprachen, Nos. 1-338.

I have taken the triliteral roots because the biliteral roots are too general and vague, and the quadriliteral too modern, having largely been developed after the separation of the Aryans and Finns. They are properly stems rather than roots.

The resemblance, nay, the identity is most surprising. Every one of these 18 triliteral Aryan roots in k is also found in Finnic with the same meaning. It is perfectly impossible that the resemblance in so many cases can be accidental. And they cannot be loan words, as they extend to the Asiatic languages of the Finnic class, as well as the European languages which are in contact with Aryan languages. They belong, therefore, to the Finnic Ursprache.

Comparison of Verbal Roots in Aryan and Finnic.

Kivog, a crab; Latin cor-nu; English | iron, gör, a plough. horn, hart.