Interesting, however, as are such observations on the animal stage of sound-production in the human infant, they do not touch the crucial period in the development of language. Mr. Buckman, indeed, regards as a remarkably dogmatic assertion Professor Max Müller’s dictum that “the one great barrier between the brute and man is language;” and he tells us that “there are more than twelve different words in the language of fowls,” on which assertion, in turn, the distinguished linguist whom he criticizes might have something piquant to say. No doubt the difference of opinion turns on the definition of the word “language.” But if, as is now generally accepted, the sentence and not the word is the distinguishing unit in language, and the copula in some form, explicit or implicit, is the pivot of the sentence, the wisest hen is probably incapable of language. The word becomes an element in language--a word proper--only when it assumes the office of a part of speech, that is to say, a constituent element in an interrelated whole. The animal “word,” if we like so to term it, is an isolated brick; a dozen, or even a couple of hundred such bricks do not constitute a building. Language, properly so called, is the builded structure, be it a palace or only a cottage; hen language, or monkey language, is, at best, so far as we at present have evidence, an unfashioned heap of bricks. It is just because language is the expression of a portion of a scheme of thought that it indicates in the speaker the possession of a rational soul, capable of perceiving and symbolizing the relationships of things as reflected in thought. Herein lies the practical value, for human advance in mental development, of language as a means of descriptive intercommunication. It renders explicit relationships otherwise merely implicit, and forces them to the front; and since these relationships are the stuff of which knowledge is built--without the realization of which any complex ideal scheme is impossible of attainment--the importance of descriptive intercommunication can scarcely be overestimated. And though there is no conclusive evidence of its occurrence among animals, yet we have in them the instinctive and intelligent basis on which in due course of evolution it may be securely based. III.--SOCIAL COMMUNITIES OF BEES AND ANTS