If we accept Mr. Medlicott’s interpretation as in the main correct, we have in this case: (1) common action in social behaviour, (2) community of emotional state, and (3) the suggestion of natural enemies not unfamiliar in the experience of the herd. Under uniform conditions of experience the alarm-notes of some birds may well call up, re-presentatively, salient features in previous situations. Unquestionably, in the parrot, the word-sounds they imitate become associated with definite objects of sense-experience. In the following case, a particular sound appeared to be suggestive of a particular sense-idea in the dog. The parent blackbirds, which built near a house in Clifton, were wont to give the alarm-note when marauding cats appeared in sight. This sound, it would seem, became definitely associated, in the experience of a terrier, with the animals the presence of which called it forth; and on hearing the alarm note the dog would rush out into the garden, apparently, as I am informed by his mistress, in fullest expectation of a pleasant worry. It is a not improbable hypothesis, therefore, that in the course of evolution the initial value of uttered sounds is emotional; but that on this may be grafted in further development the indication of particular enemies. If, for example, the cry which prompts instant flight among the pigs is called forth by a tiger, it is reasonable to suppose that this cry would give rise to a representative generic image of that animal having its influence on the conscious situation. But if the second cry, for defence, was prompted sometimes by a leopard and sometimes by some other minor foe, then this cry would not give rise to a re-presentative image of the same definiteness. Whether animals have the power of intentionally differentiating the sounds they make to indicate different objects, is extremely doubtful. Can a dog bark in different tones to indicate “cat” or “rat,” as the case may be? Probably not. It may, however, be asked why, if a pig may squeak differently, and thus, perhaps, incidently indicate on the one hand “tiger” and on the other hand “leopard,” should not a dog bark differently, and thus indicate appropriately “cat” or “rat”? Because it is assumed that the two different cries in the pig are the instinctive expression of two different emotional states, and Mr. Medlicott could distinguish them; whereas, in the case of the dog, we can distinguish no difference between his barking in the one case and the other, nor do the emotional states appear to be differentiated. Of course, there may be differences which we have failed to detect. What may be regarded, however, as improbable, is the _intentional_ differentiation of sounds by barking in different tones with the _purpose_ of indicating “cat” or “rat.” Mr. R. L. Garner, in a work[83] which unfortunately contains much hasty and immature generalization, distinguished nine sounds made by capuchins. But none of these, so far as can be gathered from