420 D'A. w. THOMPSON: REGENERATION OF LOST PARTS, ETC. lost part is but a recurrence to embryonic processes, a re-travelling of a familiar road : and not an independent present effort of the cells to mould themselves in conformity with function an im- mediate manifestation of reciprocal interaction between the tissues and their surroundings. And the already described instances of retrogressive tendencies point to the same conclusion. In all the processes of growth and evolution as well as of repair and regeneration, we are forced to acknowledge that the cell is influenced by something akin to habit or custom, by some tendency to do as its ancestors have done before ; and the most complex cell-divisions and most wonderful fabrications of organic mechanism seem but the indication of perpetual bondage to an eternal routine. And many of the phenomena of regeneration, of budding, and of parthenogenetic reproduction seem to indicate that it is but a sort of accident, not the defect of a peculiar attribute, that prevents any single cell of an animal from repro- ducing the entire organism. I will give one concrete instance of what I mean by the ten- dency of growth to persist in accustomed and long previously determined lines. I believe that the forms of the surfaces of joints are determined by strictly mechanical causes ; that if our limb-bones were only roughly blocked out in soft cartilage, their extremities would tend in the course of action to assume the shape that they actually possess ; convexities and concavities would be modelled out where we actually find them. But in the foetus in utero the motionless limbs acquire passively their proper conformation ; and they do so because a certain stereotyped behaviour has been stamped into their tissues ever since the time when that behaviour was directly influenced by mechanical forces. Mr. Haldane, however, would probably deny that all this really affects the validity of his speculations ; and very likely it only limits their range. For though we bring in with some show of proof the hypothesis of old habit to account for the cells dividing in such a manner as to reproduce the newt's lost limb, yet some- thing further is needed to account for its accurate regeneration when the line of section is oblique or irregular. We cannot but see that the procedure of each cell is regulated and conditioned by the needs of all the rest. Whatever histological or proto- plasmic continuity there may be between cell and cell, we may dimly fancy some system of forces interconnecting them, which dynamical system may be again reciprocally influenced by the conditions of the outer world.
