When a fragment of an organism transforms, under appropriate conditions, into a typical individual, the process includes degenerative aa well as regenerative phases. There is always some simplification of the structures present, whose character and amount is determined by the degree of specialization which has been attained. The smaller the piece, within certain limits, and the younger physiologically, the more nearly does it return to embryonic conditions, a fact which can be studied admirably in the hydroid Corymorpha. In some cases the simplification is accomplished by abrupt sacrifice of highly specialized parts, as in Corymorpha, when in a process of simplification connected with acclimatization to aquarium conditions, the large tentacles of well-grown specimens fall away completely from their bases. In other hydroids (e. g., Campanularia) the tentacles may be completely absorbed into the body of the hydranth from which they originally sprang. Among tissue cells degenerative changes may be abrupt, as in the sacrifice of the highly specialized fibrillae in muscle cells; or they may be very gradual, as in the transformation of cells of one sort into another that occurs in the regeneration of tentacles in Tubularia. An interesting case of absorption of parts came to my notice while studying the larvae of the pennatulid coral Renilla some fifteen years ago. As will be remembered, Renilla possesses eight tentacles with numerous processes pinnately arranged. During a period of enforced starvation, these pinnae were gradually absorbed, and the tentacles shortened, from tip to base. With the advent of food--in the form of annelid eggs--the reverse of these events took place. The tentacles lengthened and the pinnae reappeared, the larvae assuming their normal aspect. It appears, then, that in some circumstances at least, the process of simplification may resemble very nearly, even in details, a reversal of the process of differentiation. That one is actually in every respect the reverse of the other is undoubtedly not true. This, however, is not to be wondered at. Mechanical inhibitions that are so conspicuous in some cases (e. g., Corymorpha) are to be expected to a certain degree in all. The regenerative process itself depends upon the cooperation of many physical and chemical factors, in many and complex physicochemical systems in varying conditions of equilibrium. And it is important to note that even the equilibrium reactions by which a single proteid in the presence of an enzyme, is made and unmade, do not appear always to follow identically the same path in opposite directions.[3] [3] Robertson, vid. sup., p. 269.