Here we find ourselves in the presence of two kinds of activities which at first seem very dissimilar,-those which all the representatives of a species produce without actual apprenticeship, and those which are acquired by each individual in the course of its own life. The first include all those actions which are the result of tropisms, of organic and of species memory. They are not due to any intention of the animal and present themselves as reflexes, but as reflexes which are coordinated so as to put the whole organism into action. They belong in the domain of instincts. The second appear very different, since they imply choice and presuppose a certain individual intention ; but we have seen that these very quickly take on the character of habits, and even turn into automatic actions. If these individual habits can be transmitted by heredity they will become similar to the first, and we shall find ourselves prepared to comprehend the problem of the psychology of insects, to know the mechanism by which these often extremely complex instincts were acquired, have evolved, and still continue to evolve. The question is highly important, and we will try to solve it in the light of facts, but before beginning we should know whether or not insects are capable of modifying their habits.

Can insects, spontaneously modify their habits? Animal psychology is a very delicate science, because more than any other it risks the confounding of the objective with the subjective. The intimate phenomena of animal behavior escape direct observation, and it is only from the manner in which they manifest themselves that we are able to understand them. Then comes in the personal judgment of the observer, and two things present themselves : on the one hand, reasoning by analogy it gives the animals human faculties (an-thropocentrism) ; on the other it shows an excessive tendency to follow from the bottom to the top the stair of living beings and to reduce all the animal activities of the lower organisms to purely mechanical manifestations.

What shipwrecks have been caused by these rocks, even among the most highly esteemed biologists! Against the first most of the biologists have struck, from Reaumur to Lubbock; among others, a disciple of Darwin, Romanes, who has popularized anthropocentrism in his remarkable work on "The Intelligence of Animals." Against the second rock the savants of the German mechanistic school have dashed themselves, •-Wundt, who denies all psychic faculty in animals, even in the most advanced ; Uexhull, who declares that '1 For the biologist, animal psychology cannot exist"; Bethe, who accords, however, certain psychic aptitudes to vertebrate animals. In his book on "Vision," Nuel ranges himself among the most advanced of the scholars of this school who have revived under a new scientific aspect, although with strange scientific terms, the animal mechanism of Descartes.

Shall we give some examples of the diametrically opposite tendencies of these two schools? We have only to choose from the works in which Lubbock and Bethe have accumulated their observations upon and their experiences with the social Hymenoptera, notably the ants and the bees. Lubbock observes that "the anthropoid apes, in their structure, evidently approach man more nearly than any other animal," while the ants can "justly pretend" to the same rank "by the degree of their intelligence." Such, however, is not the opinion of Bethe (1898). "Should we attribute psychic qualities to the ants and to the bees?" asks the Strasburg professor, who, after having experimented and observed, replies to this question in the following manner: "It seems that the whole group of invertebrate animals lead a purely reflex life. . . and that these animals perform as simple machines all the acts which appear so intelligent. ' '

So the two methods of reasoning (the one descending from man to the inferior animals, the other mounting toward man, starting with the simplest organisms) bring diametrically opposed conclusions. "That was to be foreseen," says Claparède1; "the descending method, as a matter of fact, speaks only in terms of consciousness; the ascending method knows only the mechano-physiological tongue." Our embarrassment would be great if we were bound to follow the one school or the other, but this is not the case, and we will wisely stay in the domain of observation and experiment. But it is not to be doubted that an animal can show proof of psychic aptitudes, of understanding, when it is "in condition to learn and to modify its comportment," its hereditary habits. That is the decision of Bethe. It was equally, I think, that of Loeb and of Bohn when they coined the expression "associative memory," although this memory becomes very quickly, at least with the insects, automatism.

1 E. Claparède, Les Animaux, sont-ils conscients? 1901. (Rev. Phil., Paris, v. 26. pp. 481-498).

We are here in the presence of a method which promises to recognize psychic actions without rising or descending in the animal scale. It suffices to follow it in making an abstract theory, leaving to the facts the signification which belongs to them. To speak strictly, in studying the variation of habits of the Osmia bees, of the Pompilus wasps, the cockroaches, the crawfish, and the crabs, we have already concluded that these animals are capable of modifying their habits and of learning to take on new ones. But this was under experimental conditions which were more or less different from those which are normal to the organism. To avoid this criticism, let us examine the insect left to itself in the surroundings among which it passes its life.