While this process of fusion is being traced in the case of substantives and adjectives, it becomes at the same time observable that the definition of verbs is gradually growing more and more vague, until it is difficult, and eventually impossible, to distinguish a verb at all as a separate part of speech. Thus we are led back by continuous stages, or through greater and greater simplifications of language-structure, to a state of things where words present what naturalists might term so generalized a type as to include, each within itself, all the functions that afterwards severally devolve upon different parts of speech. Like those animalcules which are at the same time but single cells and entire organisms, these are at the same time single words and independent sentences. Moreover, as in the one case there is life, in the other case there is meaning; but the meaning, like the life, is vague and unevolved: the sentence is an organism without organs, and is generalized only in the sense that it is protoplasmic. In view of these facts (which, be it observed, are furnished by languages still existing, as well as by the philological record of languages long since extinct) it is impossible to withhold assent from the now universal doctrine of philologists—“language diminishes the farther we look back in such a way, that we cannot forbear concluding it must once have had no existence at all.”[250] * * * * * From all the evidence which has now been presented showing that aboriginally words were sentences, it follows that aboriginally there can have been no distinction between terms and propositions. Nevertheless, although this follows deductively from the general truth in question, it is desirable that we should study in more detail the special application of the principle to the case of formal predication, seeing that, as so often previously remarked, this is the place where my opponents have taken their stand. The reader will remember that I have already disposed of their assertions with regard to the copula. It will now be my object to show that their analysis is equally erroneous where it is concerned with both the other elements of which a formal proposition consists. Not having taken the trouble to acquaint themselves with the results of linguistic research, and therefore relying only on what may be termed the accidents of language as these happen to occur in the Aryan branch of the great language-tree, these writers assume that a proposition must always and everywhere have been thrown into the precisely finished form in which it was analyzed by Aristotle. As a matter of fact, however, it is now well known that such is not the case; that the form of predication as we have it in our European languages has been the outcome of a prolonged course of evolution; and that in its most primitive stage, or in the earliest stage which happens to have been preserved in the palæontology of