We are thinking of the general and important question, what types of adaptations may be expected in the field of physiology and whether there may be certain classes of regulatory events which possibly might be expected to occur in the organism on *a priori* grounds, but which, nevertheless, are to be regarded as impossible after a more intimate analysis of its nature, even at the very beginning? Or, in other words, to what kinds of changes of the medium will an organism be found able or unable to adapt itself? We know that the *state of functioning* must be altered in order to call forth any sort of adaptation at all. Now, there can be no doubt that *a priori* it would seem to be very useful for the organism, if it never would let enter into its blood, lymph, etc., be it through the skin or through the intestine, any chemical compound that would prove to be a poison afterwards. In fact, a man, judging on the principle of the general usefulness of all the phenomena of the living, might suppose that there would exist a sort of adaptation against all poisons to the extent that they would never be allowed to enter the real interior of the body. We know that such reasoning would be incorrect. But we also can understand, I suppose, that an *a priori* analysis of a more careful kind would have reasoned differently. How could the functional state of the organism be changed, and how, therefore, could adaptation be called forth by any factor of the medium which had not yet entered the organism, but was only about to enter it? Not at all therefore is such a regulation to be expected as we have sketched; if there is to be any adaptation to poisons, it only can occur after the poison has really acted in some way, and in this case we shall indeed find regulations. You may perhaps regard this discussion as a little too academical and hair-splitting, but here again it was for the sake of ensuring a perfectly sound foundation of our chief principles that I undertook it. Very often, indeed, the question has been raised by the defenders of a mechanistic theory of life, Why then did the organisms not reject all poisons from the very beginning? We now may reply to that only--how *could* they do so? How could they “know” what is a poison and what is not, unless they had experienced it?--if we are allowed for a moment to use very anthropomorphistic language. We repeat, therefore, that the functional conditions of the organism must have been actually changed in order that an adaptation may occur. Nothing is more essential to a clear understanding of our problems than to keep fully in mind the exact sense of this definition. ON CERTAIN GROUPS OF PRIMARY PHYSIOLOGICAL ADAPTATIONS