166 MENDEL'S LAWS 



by a period of above sixty years, were lying latent in this hen-bird, 

 ready to be evolved as soon as her ovaria became diseased." 1 

 Here there can have occurred no reunion of previously divorced 

 factors and determiners. The only conceivable interpretation of the 

 facts is that the ancestral characters were latent', and if they are 

 latent in the pure-bred varieties, we have no right to assume that 

 they segregate in the mongrel. " Entia non sunt multiplicanda 

 prceter necessitate" 



281. Very obviously Cudnot and his followers have not taken 

 all the evidence into account. His hypothesis merely complicates 

 the problem and confuses the issues by introducing the entirely 

 gratuitous assumption that the production of characters depends, 

 not on a single unit or hereditary tendency in the germ-plasm, but 

 on two, a factor and a determiner. It must be understood that 

 there is no real evidence in favour of this hypothesis. It was 

 invented only because without some such supposition the reproduc- 

 tion of ancestral traits rendered the hypothesis of segregation 

 untenable. It is unnecessary, therefore, to consider what have 

 been euphoniously termed the ' later refinements ' of the Men- 

 delian hypothesis, such as negative ' elements ' in which apparently 

 * absence ' constitutes the alternative allelomorph. As I say the 

 reproduction of ancestral traits by pure-bred races constitutes 

 decisive proof that the Mendelian phenomena are due to alternate 

 patency and latency, not to segregation. 



282. De Vries offers yet another explanation of Mendelian 

 phenomena. 2 According to him a new species differs from the 

 parent stock by the addition of a fresh progressive character (a 

 mutation) to the sum-total of ancestral characters. A new variety, 

 on the other hand, differs from the parent stock in that it has 

 latent or patent a character which is in the opposite condition in 

 the parent stock. When, therefore, species cross they tend to be 

 more or less sterile, for the character is unpaired ; or if the union 

 is fertile there tends to be more or less of blending. When 

 varieties cross with the parent type the patent character is dominant 

 over the latent character, and there is Mendelian inheritance, with 

 segregation of patent and latent units in succeeding generations. 

 But, if this theory be correct, the albino rabbit must have been 

 latent in the yellow-grey, the yellow-grey in the albino, and the 

 black, wild-grey, and Dutch colorations in both. The ancestors 

 of both races, therefore^ must have been all these colours in turn 



1 Animals and, Plants, vol. ii. p. 29. 



1 Species and Varieties, Their Origin by Mutation. Lectures IX and X. 





