176 CHARLES DARWIN. 



are usually intermediate between the parents, but in 

 the next generation thel-e is commonly reversion to 

 one or both parent-forms, or even to a more remote 

 ancestor. The existence of abundant hybridised 

 gemmules is shown by the propagation of the cross in 

 a true form by means of bads ; but dormant gemmules 

 from the parent-form are also present and multiply. 

 In the sexual elements of the hybrid there are both 

 pure and hybrid gemmules, and the addition of the 

 pure gemmules in one sex to those in the other 

 accounts for the reversion, especially if we assume that 

 pure " gemmules of the same nature would be 

 especially apt to combine." Partial reversion on the 

 one hand, and the reappearance of the hybrid form 

 on the other, would be respectively due to a combina- 

 tion of pure with hybrid gemmules, and of the hybrid 

 gemmules from both parent hybrids. 



When characters which do not blend exist in the 

 parents, crossing may result in an insufficiency of 

 gemmules from the male alone and from the female 

 alone, and then dormant ancestral gemmules might 

 have the opportunity of development, and thus cause 

 reversion. Similarly certain conditions might favour 

 the increase and development of dormant gemmules. 

 Diseases appearing in alternate generations, or gain- 

 ing strength by the intermission of a generation, 

 may be due to the increase of the gemmules in the 

 intervening time, and the same explanation may hold 

 for the sudden and irregular increase of a weakly 

 inherited modification. 



