Whether they are the result of acquired characters or come from mutations, the innate and automatic acts which we have just studied are characterized by the fact that they are always supposed to have a psychic origin. This rule applies to mutations with not less force than to the acquired characters, for the sudden appearance of a new habit carries with it some fumbling or groping, soon rectified, and the sudden appearance of a new organ brings fumblings or gropings and analogous rectifications. When an Atya is born from an Ortmannia it is ignorant of the functioning of the claws armed to the base and the long plumy hairs which terminate them, and it is necessary to acquire habits which were not those of her progenitors; and however short the apprentice stage may be, it nevertheless demands some gleams of intellect.

It is to acts of this kind that many authors confine the term instinct, usually without considering their psychic origin. Thus defined, instinctive acts, according to Cuénot, differ from tropisms, from differential reactions, and without doubt from rhythms, "in that they are more complicated and dependent to a lesser degree upon external stimuli, being above all determined by internal stimuli." These differences are real, but it will not be difficult to cite acts of this kind as simple and as dependent upon external stimuli as the reactions due to a species memory. And if one considers that these last come from an undeniable psychism, although far back, the barrier totally disappears which separates instinctive acts, thus understood, from phenomena of differential sensibility.

In fact, we believe it is preferable to designate as instincts all the innate and automatic acts of an organism which is sound and whole: the tropisms and the rhythms which are simple reactions of riving matter, without apparent trace of choice, -that is to say of psychism-even when they imply, like rhythms, an organic memory; the manifestations 'adapted to a differential sensibility which presuppose a species memory and consequently a very ancient psychism; finally, the acts which come from an individual or associative memory which are the automatic expressions of a higher grade of psychism. Thus instincts group themselves into different categories and are far from having, all of them, the same origin. That is why we believe it illogical to see in them the manifestations of a faculty which one opposes to intelligence, under the name instinct. This latter term is frequently employed, but it should not be forgotten that it expresses phenomena of a very différent kind.

Automatic and innate, like reflexes, instincts are closely related to reflexes, and the sole character which permits us to distinguish them is that they demand the integrity of the organism, while reflexes can be very well produced in portions of organisms immediately after their separation from the body. For a definition, we cannot have true reflexes with beings where the nervous system is not differentiated, and therefore we do not know how to join to reflexes the tropisms of the lower organisms. But it is not so with the articulates, and it may be said that with these animals instincts require the condition of age and the help of the cerebral ganglia, while reflexes are often produced in decapitated or sleeping individuals.

However, even this difference is not absolute, as is shown by hypnotized subjects, and even better by the phenomena of dreams and somnambulism. In the hypnotized subjects there is still a will which commands, that of the hypnotizer; in the somnambulist and in the subject who is dreaming, reflexes govern the activity, either according to chance, which makes one or the other act on the nervous system, or in a well-determined sense which gives the acts accomplished an appearance of reason, as one often observes with sleep-walkers. How, then, can we distinguish instincts from reflexes'?

In our opinion, in these extreme cases distinction is not possible, and the articulates give us a manifest proof. There are cases, in fact, where these animals are totally absorbed by the reflex life without any regard to contingencies. Who has not seen groups of ants excite themselves over a twig which they were trying to bend, vainly,- because they were all pulling against one another ! Certain crabs have the habit of gluing to their backs all sorts of objects, and this habit gives them the advantage of concealing themselves in the places where they live, but they stick on also strikingly noticeable objects foreign to the environment. The spiders of the genus Pardosa attach to their spinnerets the silk cocoon which contains their eggs, and carry it, but they accept with indifference any foreign body of the same form and to this inert object they devote the same care as they give their precious treasure.

We know how, after having paralyzed their victim, certain predatory Hymenoptera grasp it by the neck and squeeze it with the mandibles in order to make it disgorge drops which they a;bsorb. Ordinarily this is done carefully, in order not to kill the paralyzed victim, but it is often also very brutal and causes wounds from which the victim does not recover, so that it is a dead body, and not an inert but living prey, which the wasp puts into her cell to nourish her future larvae. Marchai has noted an instance of this sort with Cerceris ornata, which hunts for its young the solitary bees of the genus Halictus. Fabre has noted still more suggestive facts. A certain Pelopœus is a paralyzing wasp which stores up in masonry cells the spiders that are necessary for food for its larvae. Let us take a cell to which the insect has carried its first spider, upon which it has laid an egg. How will the wasp act if you take away this victim? "It acts absurdly," says Fabre. A second victim was brought, which disappeared also, to be replaced by a third, and so on. The experiment continued for two days. "At the twentieth victim, tired out by these repeated expeditions, the wasp judged that the cell was sufficiently filled, and very conscientiously closed it up containing nothing at all." The same wasp made its nest of several cells attached longitudinally, which it covered afterward with a uniform rough coating. One of these structures, which had just been completed, was entirely removed, leaving only a faint trace on the wall where it had been. On its return, without hesitation the Pelopœus alighted on the place where the nest had been and deposited there its clay pill, and it continued to make the rough coating over the absent nest. "Two days afterward," says Fabre, ' ' I visited the hidden place, and it did not differ in appearance from a completed nest."

The mason bee is not less inept in changing her work while this is under way. If she is engaged in building, one can give her in exchange an entirely completed cell filled with food, but she will keep on building and adding material to the already completed cell. If she is in the act of collecting food, she does not try to finish an incomplete cell which has been given her, but seeks rather a strange cell in the proper condition in which to store her honey, Fabre says ;

The bee that is building, and to whom I offer a cell already built, full of honey, does not renounce her mortar for that ; she is making masonry, and having started on that task, guided by an unconscious impulse, she must continue to make masonry, although her task will be useless, superfluous, contrary to her interests.

If the insect is engaged in provisioning, " she will never decide to abandon the pollen brush for the trowel."

Studying similar phenomena with a predatory wasp in Eéunion, Bordage (1912) supposes that the four successive acts of these insects-construction, provisioning, egg-laying, and the closing of the cell-are reflexes caused by the passage of the egg to different points in the oviduct, each region being the seat of a definite reflex which the insect must obey. This is very possible and perfectly rational, but, however it may be with this theory, one is obliged to admit that in the cases cited our insects have acted as pure automata by a simple play of reflexes which hold them, as is the case with somnambulists.

"With their tropisms, their rhythms, the adaptive manifestations of their differential sensibility,-above all, with their power of transforming habits into automatic actions,-the articulates are essentially animals of instinct whose activities consist principally of automatisms, but automatisms dominated by cerebral power. One can hardly see in them "simple reflex machines," for they know how to bend to circumstances, to acquire new habits, to learn and to retain, to show discernment. They are, one can say, somnambulists whose minds awaken and give proof of intellect when there is need for it. This takes us a long distance from the mechanism of which Bethe has made himself the protagonist.

1 Two essential peculiarities characterize the activity of insects : on the one hand the presence of multiple appendages which are more or less perfectly adapted; on the other, the power very quickly to transform acts which are intelligent at first into automatic acts. The second is, without doubt, a consequence of the first, for the appendages are tools scarcely fixed in their forms and in their functions, and at any rate this is undeniably the principal factor in the evolution of the articulates. Thanks to it, as a matter of fact, the automatic activity of the animal is able to enrich itself with new elements borrowed from intelligence and therefore adapted to new necessities. A substratum of activity is produced and developed which permits intelligence, as Bergson says, to push out on the wings of instinct. But it does not push out very far or very high, because its impulses are quickly fixed in an automatic form, but each time the instinctive substratum is increased so as to give the animal a larger field of activity. Thus one reaches the higher articulates, where the most 'complex automatic actions, touched with intelligence, control them, and lead to an end where they seem ruled by reason. And that is why we repeat here what we said at the beginning' of the present work: Never are the articulates so far from us a's when they appear to resemble us most.