16 DAVIS— THE TEST OF A [April 23, 



ticular crosses. De Vries ('09) has also described "triple hybrids "^ 

 when the " mutants " scintillans and lata are pollinated by such 

 species as produce the twin hybrids from Lamarckiana. In such 

 cases two of the forms have the characters of leeta and velutina 

 combined with those of the other parent, and the third form re- 

 sembles the mother, either scintillans or lata. The phenomena of 

 twin and triple hybrids is treated in detail by De Vries ('13) in 

 " Gruppenweise Artbildung." 



From a MendeUan standpoint the production of twin and triple 

 hybrids is strong evidence that Lamarckiana and such of its 

 " mutants " as behave in this manner are impure or hybrid since the 

 male or female gametes are not uniform, a point which has been 

 emphasized by several critics of the mutation theory. De Vries 

 assumes that Lamarckiana forms its different classes of gametes as 

 a result of its mutating instability but the precision of the process 

 falls completley in line with what we know of Mendelian behavior. 

 The remarkable studies of Shull show that crosses between La- 

 marckiana and cruciata give in the first generation polymorphic 

 progenies of much greater complexity than the twin hybrids of De 

 Vries. Shull's results have not been published in full but, as I 

 understand them, they indicate the interaction of several classes of 

 gametes, a condition very far from what would be expected if 

 genetically pure species had been crossed. 



Very interesting are the observations of Atkinson ('14) on first 

 generation crosses between (Enothera nutans and 0. pycnocarpa. 

 These two forms are American species recently segregated by Atkin- 

 son and Bartlett from the biennis alliance. They have bred true in 

 garden cultures. When pycnocarpa is pollinated by nutans twin 

 hybrids appear in the first generation. In the reciprocal cross 

 nutans y,pycnocarpa the same twin forms are produced and in addi- 

 tion a third type, making this generation a compound of three dis- 

 tinct forms, triple hybrids. Atkinson, apparently confident of the 

 genetic purity of nutans and pycnocarpa assumes that the determina- 

 tion of the twin and triple hybrids takes place through a diflFerential 

 division in the zygote by which factors representing certain char- 

 acters are side tracked in the suspensor cell and only those respon- 

 sible for the twins and triplets pass on to the embryo. There is no 



