President's Address. 147 



its being more or less unfertile with the parent form and 

 with other varieties in the same area. I doubt very much 

 if free intercrossing would result in the loss of all the 

 present domestic breeds. Domestic animals have strong 

 likes and dislikes, and vary greatly in their fertility and 

 prepotency. It is hence possible that, were representatives 

 of several breeds of horses, sheep, or cattle set free and 

 allowed an unlimited range, instead of giving rise to a single 

 well-marked variety they would form several varieties, some 

 of them, it may be, differing but little from the breeds 

 originally set free. So exclusive are some varieties that, 

 though confined in a limited area and allowed to intercross 

 freely, pure specimens would probably persist for an indefinite 

 period. Evidence in support of the view that new varieties 

 are not necessarily swamped by intercrossing, we have in the 

 results obtained by Mendel and his followers. 



According to Mendel's law, the offspring of two varieties 

 when interbred, yield offspring representing the two original 

 varieties as well as the crosses ; the second and succeeding 

 crosses, when interbred, yielding pure and mixed offspring as 

 before. 



Mendelian experiments may hence be said to practically 

 dispose of the assumption that intercrossing invariably results 

 in the swamping of new varieties. 



But not only are many varieties so constituted that when 

 intercrossed they sooner or later emerge in their original 

 purity; there are varieties so exclusive that when crossed 

 they invariably yield offspring having the characters of either 

 the one parent or the other, but are never intermediate. For 

 these and other reasons it follows (1) that intercrossing does 

 not inevitably lead to swamping, and (2) that infertility 

 between a variety and the parent form is not the only 

 possible means of affording a new variety the chance of 

 developing into a new species. 



We now know that there is hardly any limit to the varia- 

 tion of germ-cells. To indicate the nature of some of these 

 variations, a few instances may be given from the Equidae. 



I have an Exmoor pony that proved sterile with various 

 breeds of horses, but fertile with a zebra and a Kianf:^. 



