312 PROBLEM OF THE ORIGIN OF SPECIES. 



most decided negative answer. Because the stimuli pass through the soma 

 on their way to the germ in no wise enters into the question. That all herita- 

 ble variations hence factors in evolution arise primarily in the germ plasm, 

 and secondarily in the soma, is on the basis of all of our present evidence the 

 only acceptable hypothesis. 



The recent compilation of data by Redfield concerning "dynamic inherit- 

 ance" of speed in race horses, and of other characters in dogs and men, 

 wherein he attempts to show the inherited effects of use, fail in that there 

 exists in every case the unproven assumption that the ancestor of the race 

 was an ordinary animal, whereas the very fact that the original progenitor of 

 the American race horse was able to train so long and to such good effect 

 shows that it was in its initial make-up different from other horses, and the 

 further development of the race horse seems to be purely a selective process. 

 Were this not so there is no reason why there should not have been a dozen 

 or five hundred horses of the same class. The very fact that this original 

 race-horse progenitor was able to develop speed beyond the ordinary, and 

 keep it up longer, is about all the evidence one needs that the original progen- 

 itor varied from the common race of horses in just those characters which 

 now form the basis of the traits of American race horses. The physical 

 capacity in this case was primary, the dynamic development later, and there 

 is no reason to suppose that the original capacity for dynamic development 

 arose other than first in the germ plasm. As far as any evidence goes that 

 we can get at, Redfield's "dynamic development" is a selective process based 

 upon an initial germinal variation. 



Students of evolution are coming more and more to believe in the origin 

 in the germ plasm of all variations that are effective in evolution. Evidence 

 is constantly being accumulated, as Fischer, Standfuss, and MacDougal have 

 shown in a few cases, and as shown here in larger series, that permanent 

 variations are direct responses in the changed constitution of the germ plasm 

 to stimuli. How these are effected is not at present known, and is a subject 

 for investigation and not for anticipation. Further, that these variations 

 follow the law of trial and error is for these beetles fully shown, and it is 

 further shown that just this method would result in orthogenesis, and even 

 mutation, without the introduction of "forces," "ultimate vital units," or 

 "latencies." How these modifications are preserved is the second part of the 

 problem of the method of evolution. 



Although evolutionists differ in their opinions in regard to the extent 

 and importance of natural selection, all, with the possible exception of the 

 Eimerites, give it an important place in organic evolution. Of late there has 

 grown up, since the publication of De Vries's work, a tendency to ascribe to 

 "mutation" a far greater importance and to a considerable extent to substi- 

 tute that process for all others in evolution, even though De Vries distinctly 



