Are there animals capable of counting? G. Leroy is, I believe, the first who answered this question in the affirmative, in a passage which is worth transcribing, although it has been often quoted. “Among the various ideas which necessity adds to the experience of animals, that of number must not be overlooked. Animals count,—so much is certain; and although up to the present time their arithmetic appears weak, it may perhaps be possible to strengthen it. In countries where game is preserved, war is made upon magpies because they steal the eggs of other birds.... And in order to destroy this greedy family at a blow, game-keepers seek to destroy the mother while sitting. To do this it is necessary to build a well-screened watch-house at the foot of the tree where the nests are, and in this a man is stationed to await the return of the parent bird, but he will wait in vain if the bird has been shot at under the same circumstances before.... To deceive this suspicious bird, the plan was hit upon of sending two men into the watch-house, one of them passed on while the other remained; but the magpie counted and kept her distance. The next day three went, and again she perceived that only two withdrew. It was eventually found necessary to send five or six men to the watch-house in order to put her out of her calculation.... This phenomenon, which is repeated as many times as the attempt is made, is one of the most extraordinary instances of the sagacity of animals.” Since then the question has been repeatedly taken up. Lubbock devotes to it the three last pages of his book _The Senses of Animals_. According to his observations on the nests of birds, one egg may be taken from a nest in which there are four, but if we take away two, the bird generally deserts its nest. The solitary wasp provisions its cell with a fixed number of victims. Sand wasps are content with one. One species of _Eumenes_ prepares five victims for its young, another species ten, another fifteen, another twenty-four; but the number of the victims is always the same for the same species. How does the insect know its characteristic number?[16] An experiment, methodically conducted by Romanes, proved that a chimpanzee can count correctly as far as five, distinguishing the words which stand for one, two, three, four, five, and at command deliver the number of straws requested of her.[17] Although the observations on this point are not yet sufficiently varied and extended to enable us to speak of them as we should wish, it must be remarked that the cases cited are not alike, and that it would be illegitimate to reduce them all to one and the same psychological mechanism. 1. The case of insects is the most embarrassing. It is but candid to state a _non liquet_, since to attribute their achievements to unconscious numeration, or to some special equivalent instinct, is tantamount to saying nothing. Besides, we are not concerned with anything relating to instinct.