146 EVOLUTION AND GENETICS 



process can scarcely be doubted. We may not be 

 able to give the cause of a new variation, but we find 

 nothing in the occurrence of a change that produces 

 a variation that is inconsistent with chemical or 

 structural alterations in the germ material. If this 

 much be conceded, the problem of self-construction 

 of even a complicated piece of mechanism is not be- 

 yond our comprehension. At any rate the problem is 

 obviously different in kind from that of constructing 

 a mechanism whose materials do not possess these 

 properties. So long as the processes of division and 

 growth take place faster than the process of acci- 

 dental destruction or death, the living material can 

 maintain itself indefinitely. The stability of such an 

 organism is no greater, of course, than that of the 

 chemical material of which it is composed. If this 

 changes in those parts that have the property of 

 division and growth (without affecting these prop- 

 erties) something new will result, a new type, and 

 if this is able to maintain itself we can imagine at 

 least something new may be established. It is not 

 necessary to suppose that all changes will have a 

 survival value, but only that some of them may. The 

 alteration may bring the organism into a new relation 

 with its environment, or through competition with 

 the old type replace it, or it may make it possible for 

 the new type to move into an environment different 

 from that of the original type and hence escape com- 



