But however this be, Dean Byrne is disposed to regard the alternating energetic utterance of the hard, and indolent utterance of the soft vowel series, as an expression of the alternating active and lethargic temperament of the race, such alternations being themselves due to the climatic conditions of their environment. "Certainly the life of the great nomadic races involves a twofold experience of this kind, as they must during their abundant summer provide for their rigorous winter, when little can be done. Their character, too, involves a striking combination of intermittent indolence and energy; and it is very remarkable that this distinction of roots is peculiar to the languages spoken originally where this great distinction of seasons exists. The fact that the distinction [between hard and soft] is imparted to all the suffixes of a root proves that the radical characteristic which it expresses is thought with these; and consequently that the radical idea is retained in the consciousness while these are added to it[630]." This is a highly characteristic instance of the methods followed by Dean Byrne in his ingenious but hopeless attempt to explain the subtle structure of speech by the still more subtle temperament of the speaker, taken in connection with the alternating nature of the climate. The feature in question cannot be due to such alternation of mood and climate, because it is persistent throughout all seasons, while the hard and soft elements occur simultaneously, one might say, promiscuously, in conversation under all mental states of those conversing. The true explanation is given by Schleicher, who points out that progressive vocal assimilation is the necessary result of agglutination, which by this means binds together the idea and its relations in their outward expression, just as they are already inseparately associated in the mind of the speaker. Hence it is that such assonance is not confined to the Ural-Altaic group, analogous processes occurring at certain stages of their growth in all forms of speech, as in Wolof, Zulu-Xosa, Celtic (expressed by the formula of Irish grammarians: "broad to broad, slender to slender"), and even in Latin, as in such vocalic concordance as: _annus, perennis_; _ars, iners_; _lego, diligo_. In these examples the root vowel is influenced by that of the prefix, while in the Mongolo-Turki family the root vowel, coming first, is unchangeable, but, as explained, influences the vowels of the postfixes, the phonetic principle being the same in both systems.