INHERITANCE 109 



plains such instances of Reversion as the production 

 of a pigeon practically identical with the blue Rock 

 from a Barb-Fantail and a Barb-Spot as due to 

 the development of ancestral blue Rock gemmules, 

 which must be supposed to be present in all 

 pigeons in sufficient numbers to produce fully 

 formed offspring, though they usually remain in 

 a latent condition. Considering the enormous 

 number of generations that must have intervened 

 between the original ancestral blue Rock and the 

 present Barb or Fantail, and that each member of 

 each of these generations must be supposed to 

 have possessed these ancestral germs in sufficient 

 numbers to cause Reversion if an opportunity 

 occurred, the magnitude of the operation and the 

 numbers of such germs originally present become 

 simply inconceivable. 



A further difficulty is found in the consideration 

 that Pangenesis, involving as we have seen the 

 presence of gemmules corresponding to the different 

 periods of life of the parent, fails altogether to 

 explain the inheritance of the characters of old age, 

 or of any period beyond that at which the ova or 

 spermatozoa were discharged from the parent. It 

 is hardly sufficient to say in answer to this that " in 

 all the changes of structure which regularly super- 

 vene in old age, we probably see the effects of 

 deteriorated growth, and not of true development," 

 for the objection may apply not merely to the 

 period of old age, but to three-fourths or more of the 

 entire life of the animal. 



One further objection may be alluded to : On 



