72 MENDELISM 



to act, even though it was this parent which bore the seed, for the 

 same result occurred whichever was the seed bearer. This rirst 

 family of "dominants" being self-fertilised, the next remarkable 

 result was that from their seeds arose two kinds of plants, viz., 

 dominants and recessives, like the grand-parents, but always in the 

 proportion of three dominants to one recessive when the average 

 of a large number was taken. Here then was an obvious law, and 

 long series of crossings and intercross! ngs proved that the characters, 

 "dominant" or "recessive," never intermingled absolutely, but only 

 temporarily ; so that when another period of reproduction occurred 

 they were capable of dissociation or segregation. 



"In the reproductive cells," says DR. R. H. LOCK, "the domi- 

 nant and recessive characters separate completely only on being 

 represented in each cell. Such cells occur in equal numbers. 

 When the reproductive cells unite they do so at random, so that in 

 the long run the combination DD, DR, RD, and RR are equally 

 recessrve. DD is a pure dominant like the dominant parent, and 

 will always breed true to this character unless again crossed ; the 

 same applies to RR, DR and RD which are crosses like the original 

 cross ; like it they are dominant in appearance, and their further 

 behaviour is exactly the same. The apparent 3 to 1 ratio is thus 

 explained. It is really a ratio of 1 : 2 : 1. i.e., ID : 2 DR : 1 R. 

 D and R breed true. DR makes up again like the original cross." 

 In MENDEL'S experiments the same ratios of "dominant" and 

 "recessive" came out in all cases where sufficiently distinct 

 characters presented themselves; and, above all, the same "dominant" 

 prevalence in the first family was found to assert itself. This, for 

 the plant-breeder, was in itself an invaluable discovery, since with- 

 out this knowledge it often took years to establish a really reliable 

 strain. He now knows that, instead of rejecting his crosses as 

 failures, he has in them probably precisely the cross he desires, 

 and has only to cross them again to obtain it in the succeeding 

 family. Furthermore, he now knows exactly how to proceed in 

 order to obtain a pure and constant strain on systematic lines. To 

 those who may desire to pursue experiments on Menclelian lines, it 

 is advisable to point out that it is only by doing so to a fairly large 

 extent that this ratio becomes clear on averaging the results. The 

 principle underlying the predominance of the dominant characters 

 over the recessive in the results obtained has been compared to 

 that shown by mixing up a large number of black and white shots 

 in a bowl, three black and one white, and employing a blind man 



