Control of Heredity: Eugenics 277 



maintaining that his results may be explained by assum- 

 ing that Oenotjiera is a hybrid and that the various "mutations" 

 which he has described are due to segregation and recombi- 

 nation of old factors rather than to the appearance of new 

 ones. Indeed Davis has made up an Oenothera by hybridiza- 

 tion that is similar to O. lamarckiana. There are many things 

 which s'eem to indicate that this species arid probably other species 

 of Oenothera are not genetically pure and it is probable that some 

 of deVries' results may be due to this fact. 



However, if 0. lamarckiana were an ordinary hybrid it should 

 show phenomena of Mendelian splitting with the usual Mendelian 

 ratios. The fact that the "mutations" observed by deVries occur 

 only very rarely shows that they are not the ordinary emergence 

 of Mendelian recessives. But the discovery of lethal factors and 

 of "crossing over" in Drosophila has pointed out a possible method 

 of accounting for many Oenoihera "mutations." In Drosophila 

 it is known that when homologous lethal factors are received from 

 both parents the zygote is non-viable; however, if the lethal factor 

 received from one parent is balanced by a normal one from the 

 other parent the zygote is viable; in other words, all individuals 

 that are homozygous for any lethal disappear, while only those 

 that are heterozygous survive. 



If recessive factors are linked with a lethal they can not come 

 to expression, for recessives appear only when mated with other 

 recessives. But if "crossing over" should take place in such a way 

 as to break the linkage between the lethal and the recessive fac- 

 tors, the latter would, when homozygous, come to expression as 

 ordinary Mendelian recessives. 



More than half of the mutants of O. lamarckiana which were 

 first described by deVries are probably due to this cause. Shull's 

 very extensive work on this species shows that it is a persistent 

 heterozygote (hybrid) in which certain recessive characters do 

 not appear as in ordinary Mendelian segregation, because the fac- 

 tors for these characters are linked with a lethal factor and it is 



