68 APPLICATION" OF PRINCIPLES OF HEREDITY TO BREEDING. 



III. LATENCY DUE TO HOMOZYGOSIS. 



This group of facts might perhaps be better described as '^patency 

 due to heterozygosis." It includes those cases where a character 

 is patent only in heterozygotes. The following discussion differs 

 in some details from that given by Shull. 



In Shull's bean crosses, mottled first-generation hybrids occurred 

 between varieties neither of which was mottled, and subsequent 

 investigation showed that the mottling only occurred in heterozygotes. 

 Tschermak and Locke have both reported similar cases; also Professor 

 Emerson, of the University of Nebraska. In all these cases the 

 mottled beans produce progeny half of which is mottled and the 

 other half not, thus giving a departure from the usual 3:1 ratio found 

 in ordinary Mendelian characters. 



The following explanation of all the above cases is here offered. 

 The explanation vAll be given for mottled character, from which the 

 explanation of the other cases may be easily inferred. The fact 

 that in beans there are mottled varieties which breed true and 

 w^hich when crossed with the new type of mottled beans give ordinary 

 Mendelian phenomena greatly strengthens the hypothesis here stated. 

 Let us suppose that originally the mottling was due to two correlated 

 characters; that is, to two characters which are always transmitted 

 together. We may represent the determiner for this double character 

 as M". The formula for those strains of mottled beans which re- 

 produce true to type would then be M M . If we suppose that in 

 some strains of these mottled beans the character a becomes latent 

 or is lost, while in other strains the character c does likewise, while 

 in still others both a and c become latent, we get three types of 

 nonmottled beans, the formulae for which are M^M^', MgMc, and MM. 

 A cross between the first and second of these nonmottled types would 

 give mottling of the character found by Shull. This cross would have 

 the formula M^^M^. In the next generation this would split up into one- 

 fourth M«M«, one-half M^M^, and one-fourth M^M^, in which only the 

 heterozygotes would be mottled, for it is only in them that we have 

 both factors of the mottling. Either of the three types of nonmottled 

 beans crossed with permanently mottled beans would give the ordi- 

 nary Mendelian behavior of the mottled character in which in the 

 second generation we should have three mottled to one nonmottled. 



This same explanation is in accord with the facts in the case of the 

 blue color in Andalusian fowls and the purple color in Imperial prim- 

 roses. In the case of the blue Andalusians the blue may not be a rever 

 sion to a lost character, but may be, in a sense, a new character; but on 

 the above explanation its presence in heterozygotes is assumed to 

 be explicable on exactly the same basis as the presence of mottling 

 in those beans in which it occurs only in heterozygotes. 



165 



