ACQUIRED CHARACTERS 171 



thev are transmitted in the same wav, so that, at 

 best, the special environment inducing them can only 

 be said to make their occurrence more frequent. 



Finally there is a considerable body of evidence 

 showing that characters, whose development is 

 known to be affected by environmental influences 

 (which therefore might be supposed to be the very 

 best kind of material to exhibit the effect of acquired 

 characters) are not affected by the changes induced 

 in their parents by the environment. There are sev- 

 eral striking cases of this kind that have been met 

 with in the course of our experiments with vinegar 

 flies. There is a race of these flies that have been 

 long inbred, in order to make them uniform in a 

 genetic sense, in w^hich the eves are entirely absent 

 in most individuals, but occasionally one or both 

 eyes may be present much reduced in size. If the 

 flies that have these small eves are bred to each other 

 they arive exactly the same results as when their eye- 

 less brothers and sisters are bred together. As each 

 stock culture gets older, more and more of the flies 

 that emerge have eyes, and, towards the end, an in- 

 creased number of the flies have both eyes present 

 and almost full size. If some of these are used as the 

 parents of a new generation, the results obtained are 

 precisely the same as when eyeless flies are used. 

 What better evidence could we hope to obtain to 

 show that the presence of a character in the individ- 

 ual has no influence on the reproductive cells i This 



