366 ACTION OF NATURAL SELECTION 



may produce more obvious degenerative changes, 

 which will of course continue and be increased 

 during extra-uterine growth. This second genera- 

 tion of dogs, besides being modified in external 

 characters, will therefore have the nature of their 

 internal secretions more altered than had the first 

 generation. These changes will react still further on 

 their offspring during intra-uterine development, and 

 BO on. 



Our conclusions as to the reaction of the germ-plasm 

 to the external conditions of environment place a much 

 higher value on somatic variations as a factor in Evolu- 

 tion than that accepted by Weismann and his followers. 

 It is for this reason that the effects of environment in 

 the production of variations have been dealt with at 

 such length in the preceding chapters. Every obvious 

 effect produced in an organism by the direct action of 

 the environment, may, in my opinion, be accompanied 

 by a more or less corresponding, though much slighter, 

 effect upon the determinants in the germ-plasm, and 

 express itself in the next generation as an apparently 

 cumulative effect of the changed environment. How 

 often this possible influence on the germ-plasm actually 

 shows itself, and what may be the numerical measure 

 of its extent, can only be determined by long continued 

 observation and experiment. 



As we shall see in the next chapter, somatic varia- 

 tions may be of very great importance in evolution by 

 reason of their adaptiveness to sudden changes of en- 

 vironment; but, quite apart from any question of adap- 

 tation, it is probable that they may be of value in 

 affording Natural Selection a better chance of exert- 



