An insect is extraordinarily stupid and inadaptable to all things not related to its instincts. Nevertheless I succeeded in teaching a water-beetle (_Dytiscus marginalis_) which in nature feeds only in the water, to eat on my table. While thus feeding, it always executed a clumsy flexor-movement with its fore-legs which brought it over on its back. The insect learned to keep on feeding while on its back, but it would not dispense with this movement, which is adapted to feeding in the water. On the other hand, it always attempted to leap out of the water (no longer fleeing to the bottom of the vessel) when I entered the room, and nibbled at the tip of my finger in the most familiar manner. Now these are certainly plastic variations of instinct. In a similar manner some large Algerian ants which I transplanted to Zurich, learned during the course of the summer months to close the entrance of their nest with pellets of earth, because they were being persecuted and annoyed by our little _Lasius niger_. In Algiers I always saw the nest-opening wide open. There are many similar examples which go to show that these tiny animals can utilise some few of their experiences even when this requires a departure from the usual instincts. That ants, bees, and wasps are able to exchange communications that are understood, and that they do not merely titillate one another with their antennæ as Bethe maintains, has been demonstrated in so many hundred instances, that it is unnecessary to waste many words on this subject. The observations of a single predatory expedition of Polyergus, with a standing still of the whole army and a seeking for the lost trail, is proof sufficient of the above statement. But, of course, this is not language in the human sense! There are no abstract concepts corresponding to the signs. We are here concerned only with hereditary, instinctively automatic signs. The same is true of their comprehension (pushing with the head, rushing at one another with wide-open mandibles, titillation with the antennæ, stridulatory movement of the abdomen, etc.). Moreover, imitation plays a great rôle. Ants, bees, etc., imitate and follow their companions. Hence it is decidedly erroneous (and in this matter Wasmann, Von Buttel, and myself are of but one opinion) to inject human thought-conception and human ratiocination into this instinct-language, as has been done to some extent, at least, even by Pierre Huber, not to mention others. It is even very doubtful whether a so-called general sensory idea (i. e., a general idea of an object, like the idea “ant,” “enemy,” “nest,” “pupa”) can arise in the emmet brain. This is hardly capable of demonstration. Undoubtedly perception and association can be carried on in a very simple way, after the manner of insects, without ever rising to such complex results. At any rate proofs of such an assumption are lacking. But what exists is surely in itself sufficiently interesting and important.