NATURE AND CAUSE OF VARIATION. 301 



of two situations. Either the whole hypothesis of ids, determinants, and bio- 

 phores is a product of the imagination, or, if they do exist, further attempts 

 to solve the riddle of evolution will be futile. 



In the experiments with Leptinotarsa new permanent variations arose as a 

 direct response to stimuli applied to the germ plasm (germ cells). In one 

 experiment, in the rise of a -five-brooded race, there was a pure, perfectly 

 constant inheritable character arising as the response to stimuli applied to the 

 germ plasm. Eleven years of study of this and related genera have shown 

 that in none of the family, or relatives of the family, are there traces of -five- 

 brooded races or species. How, then, can we explain this on the basis of the 

 id-determiinant-biophore hypothesis ? By supposing ids, determinants, or bio- 

 phores to have existed latent in the germ plasm from some extremely remote 

 -ancestor who had a five-brooded condition, and then by germinal selection 

 lost it, a few stray ids of the five-brooded condition remaining latent for 

 countless generations, finally to appear in my experiments? If we accept 

 such an explanation science has no further excuse for its existence. 



I believe that the rise of this five-brooded race is a direct experimental 

 ■demonstration of the falsity of the whole id-determinant-biophore hypothesis. 

 Possibly we could explain the origin of this race upon the supposition of 

 mutation, sudden jumps of ids or determinants from a two-brooded to a five- 

 iDrooded condition, but here we encounter the cherished hypothesis of germi- 

 nal selection, and obviously there was no time in this experiment for selection 

 to produce the results obtained. Moreover, the mutation hypothesis, with its 

 l^asic assumption of pangenes, each carrying "unit characters" absolutely iso- 

 lated from one another, afifords only a superficial explanation. The most 

 logical interpretation is that this variation is a response of the germ plasm, in 

 changed chemical or physical constitution, due to stimuli which produce an 

 inheritable variation differing in no respect in its transmissibility and perma- 

 nency from the other variations produced. 



Weismann admits that unicellular organisms may vary even in response to 

 external stimuli, which we know follow the general laws of variability, and I 

 -can see no reason for a difference in kind of organic variability. Moreover, 

 my experiments with Leptinotarsa point clearly to the unity of variability 

 in its conformity to the general law of variability expressed in the method of 

 trial, with errors distributed according to the probability of error. Now, 

 while I believe that the law of variation is a general one, it does not and 

 never can explain anything. All that we can ever get out of this law of 

 variation, or any other, for that matter, is a more and more concise formula 

 whereby we may express our accumulated, systematized experiences con- 

 cerning variations. The best we can do is to describe and determine the rela- 

 tions and sequences of stages, and in the sense that an antecedent stage is the 

 cause of a following stage, we can determine "causes," but we shall never 



