no meaning. The words which stand for substantive impressions may have suggestive value through direct association. The word “cat” or “rats” may have for the dog a very definite suggestive value; and hence some people fancy that when they say to their dog, “There is a cat in the garden,” the animal understands what they say. But it is quite sufficient to suppose that the word “cat” has suggestive force, all the rest being for the dog mere surplusage of sound. When we talk to our four-footed companions, how much can they be said to understand of what we say? Perhaps a score of words have for a dog a definitely suggestive value, each associated with some simple object or action. “Out,” “down,” “up,” “walk,” “biscuit,” “cat,” “fetch,” and so forth elicit appropriate responses. Even with these, tone is more suggestive than articulation, and in each word the salient feature is the chief guide. When I said “Whisky,” for example, to my fox-terrier, he would at once sit up and beg; not because his tastes were as depraved as those of his master, but because the _isk_ sound, common both to “Whisky” and “biscuit,” was what had for his ears the suggestive value. In a paper on the “Speech of Children,”[90] Mr. S. S. Buckman exhibits the animal stage in the incipient speech of the human infant. We cannot here discuss, still less criticize, his paper. One or two examples will serve to illustrate how instinctive sounds may serve as the basis for subsequent speech. He regards _ma_ as primarily a forcible expression of an emotional state. “If the child require attention it makes the loudest noise which it can produce; the parting of the lips and opening of the mouth to the widest extent while the full volume of breath is emitted produces the sound _ma_.” At first the sound seems to have the value of a simple expression of an emotional state. “But if the infant require attention it is its mother whom it wants, and from whom it receives this attention; therefore _ma_ very soon comes to be recognized as the call for mother, and, by a further step in development, as the name for mother.” Here, if we accept the interpretation, we have the passage from the emission of a sound as the expression of emotion to the use of the sound from its association with a particular object of sense-experience to indicate that object. Similarly, according to Mr. Buckman, with _kah_. At first “a strong sign of displeasure at anything nasty to the taste,” it passes, we are told, into a symbol for the bad; hence κακός; and is perhaps narrowed down to the particularly offensive κάκκη. _Da_ and _ta_ are regarded as recognition sounds, the former being associated eventually with the father, the latter with strangers. This appears somewhat hypothetical, but, granting the accuracy of Mr. Buckman’s interpretation, these sounds also illustrate again the transition from the expression of an emotion to sounds indicative of particular objects of experience.