154 BLASTOGENIC VARIATIONS. 



three other generations back be selected, then in .9879 

 of their full strength. That is to say, " after six gener- 

 ations of selection the selected individuals will, without 

 further selection, breed true to the selected type 

 within nearly 1 per cent, of its value." In fact, prac- 

 tically all regression towards mediocrity will have been 

 weeded out. Supposing now, some variety of a species 

 which had been bred true to its varietal characters 

 for only two generations were crossed with another 

 variety of the same species which had been bred true 

 to Us characters for six, then the resulting offspring 



8049 



would receive ' ~ = .4025 of the characters of one 

 & 



9879 

 parent, = . 4940 of those of the more thorough- 



<0 



bred parent, and .1035 of unknown blood. Knowing 

 as we do that many characters show little or no tend- 

 ency to blend, it would not be very remarkable if the 

 offspring resembled the more thoroughbred parent to 

 the partial or almost complete exclusion of the ill bred. 

 That is to say, the one parent would prove itself 

 strongly prepotent, simply through its characters hav- 

 ing become fixed through inbreeding. 



Hybridisation. Though, as a rule, intercrossing be- 

 tween different varieties of the same species tends to 

 produce uniformity of character, yet it may also very 

 frequently lead to the production of increased varia- 

 bility, not only by the partial or complete absence of 

 blending of the parental characters, but also by the ap- 

 pearance of seemingly fresh characters, due to rever- 

 sion or some other cause. 



Our knowledge of the laws governing hybridisation 



