68 



greater number of generations. But, that the act of 

 crossing, in itself, gives an impulse towards Reversion, 

 as shown by the reappearance of long-lost (!) charac- 

 ters, has never, I believe, been hitherto proved. The 

 proof lies in certain peculiarities which do not charac- 

 terize the immediate parents, and therefore cannot have 

 been derived from them, frequently appearing in the 

 offspring of two breeds when crossed, which peculiari- 

 ties never appear, or appear with extreme rarity, in 

 these same breeds, as long as they are precluded from 

 crossing." 



This remark, of his, attests strongly, that he has 

 settled in his own mind, that all the improvements 

 which arise, are due to Reversion. He says, "The 

 proof lies in " the appearance of " certain peculiarities 

 which do not characterize the immediate parents and 

 therefore cannot have been derived from them." It 

 will be observed, that this circumstance, viz., of the 

 characters not having been derived from the immedi- 

 ate parents, he makes the sole criterion of their being 

 due to Reversion. The rest of his sentence, has rela- 

 tion to his proof, that crossing gives an impulse to 

 this Reversion. That is the true law, namely that posi- 

 tive characters, which are not derived from the immedi- 

 ate parents, are due to Reversion. Had he, however, 

 formulated this rule in set terms, it would have pro- 

 voked an immediate response to his gratuitous as- 

 sumption, that there is no limit to variations. This, 

 his assumption, of no limit to improvements, is the 

 witching point in the whole controversy. Yet, strange 

 to say, such assumption is not formulated even once^ 

 but remains a tacit assumption merely, throughout all 



