Thirdly, higher up in the stratified deposits, we meet with overwhelming evidence of the connotative extension of these denotative terms. Indeed, many of these terms have probably undergone a certain amount of connotative extension as the condition to their having survived as roots; and, therefore, in these lowest deposits it is difficult to be sure that an apparently denotative term is not really a term which has undergone the earlier stages of connotative extension. If such were the case, we can understand the loss of any onomatopoetic significance which it may originally have presented. But, however this may be, there is an endless mass of evidence to prove the subsequent and continuous growth of connotative extension throughout the whole range of philological time. Lastly, as regards the predicative phase, we have seen that philology shows the same order and method to have been followed in the race as in the child. In the growing child, as we have seen, pre-conceptual predication is contemporary with—or occupies the same psychological level as—the connotative extension of denotative terms. Indeed, the very act of connotation is in itself an act of predication—if in the conceptual sphere, of conceptual predication (denomination); if in the pre-conceptual, of pre-conceptual. Again, in the psychogenesis of the child we noted how important a part is played in the development of pre-conceptual predication by the mere apposition of connotative terms—such apposition being rendered inevitable by the laws of association. If A is the connotative name for _A_, B the connotative name for _B_, when the young child sees that _A_ and _B_ occur together, the statement A B is rendered inevitable by “the logic of events;” and this statement is a pre-conceptual proposition. Now, in both these respects philology yields abundant parallels. The quotations which I have given conclusively prove that “every word must originally have been a sentence;” or, in my own terminology, a pre-conceptual proposition of precisely the same kind as that which is employed by a young child. If it be replied that the young child is without self-consciousness, while the primitive man was not without self-consciousness, this would merely be to beg the whole question on which we are engaged, and, moreover, to beg it in the teeth of every antecedent probability, as well as of every actual analogy, to which appeal can possibly be made. If it be true—and who will venture to doubt it?—that “language diminishes the further we look back, in such a way that we cannot forbear concluding it must once have had no existence at all,” will it be maintained that the man-like being who was then unable to communicate with his fellows by means of any words at all was gifted with self-consciousness? Should so absurd a statement be ventured, it would be fatal to the argument of my adversaries; for the statement would imply, either that concepts may exist without