36 J. S. HALDANE. cells, for instance, assume the form, consistence, size, and colour suited to the function which the contractile pigment- cells beneath the epidermis have to perform, while others assume the properties required for muscle-cells. Now here the conception of reciprocity fails us. For in a system of parts which reciprocally determine one another, each part, as we have seen, is independent as concerns its own individual properties. We regard the planets in the solar system, for example, as reciprocally acting on one another, but yet as each independent of the system so far as concerns such pro- perties as shape and consistence. In the case, however, of life, the fact of the determination in a certain way of the individual properties of the parts has to be explained, as well as the behaviour of the parts relatively to one another. It is common to account for the fact that the properties of any part of an organism are adapted to the function which they perform, by supposing that they have been made so by forces acting from outside of the life of the organism. These forces are found, either in a God existing independently of the physical universe, or in the environment of organisms. Thus it is argued that the parts of organisms are constantly undergoing slight modifications under the influence of the environment ; that some of these modifications give the organism which has undergone them an extra chance of survival ; and that, as offspring resembles parent, the advan- tageous modifications must necessarily tend to be transmitted and preserved in a stock, which will prevail over other stocks. This account of the matter implies, then, that the presence of design is, in ultimate analysis, accidental to the parts, in the same sense as that in which the form of a marble bust is accidental to the marble. But how can such an explanation be reconciled with the fact, for instance, that the developing cells of the new limb of a newt assume the properties suited to their future function, in whatever way the method of amputation may be varied? It is not only the behaviour, under these circumstances, of the cells relatively to one another that has to be accounted for ; but also the fact that as to its individual properties each cell constitutes itself in a certain manner according to its particular future function, and does so, not in a blind mechanical manner, but allowing for complicated disturbing conditions. This latter fact becomes altogether unintelligible if a cell be looked upon as something whose pro- perties have been passively received from without, as the parts of a machine have received their shape, &c. For this reason, among others, Hackel's ingenious theory of the nature of reproduction seems to me to be quite untenable.
