XIV.] DO VARIATIONS ARISE IN GERM ? 265 



somewhat diminished, appeared in the offspring. Seeds 

 were again taken, and in the third generation the 

 acquired characters were still discernible. The full 

 details of this and similar experiments are waiting 

 for separate publication. 



The whole philosophy of ' ' selecting the best ' ' for 

 seed, by means of which all domestic plants have been 

 so greatly ameliorated, rests upon the hereditability of 

 these characters which arise after birth ; and if the 

 gardener did not possess this power of causing like 

 plants to vary and then of perpetuating more or less 

 completely the characters which he secures, he would 

 at once quit the business, because there would no 

 longer be any reward for his efforts. Of course, the 

 Neo- Darwinians can say, upon the one hand, that all 

 the variations which the gardener secures and keeps 

 were potentially present in the germ, but they can- 

 not prove it, neither can they make any gardener 

 believe it ; or, on the other hand, they can say that 

 the new characters have somehow impressed them- 

 selves upon the germ, a proposition to which the gar- 

 dener will not object, because he does not care about 

 the form of words so long as he is not disputed in 

 the facts. Weismann admits that "climatic and other 

 external influences ' ' are capable of affecting the germ , 

 or of producing "permanent variations," after they 

 have operated "uniformly for a long period," or for 

 more than one generation. Every annual plant dies 

 at the end of the season, therefore whatever effect 

 the environment may have had upon it is lost, unless 

 the effect is preserved in the seed ; and it does not 

 matter how many generations have lived under the 

 given uniform environment, for the plant starts all 



