452 



BOTAXICAL GAZETTE 



[DECEMBER 



female gametes. So far, there is no evidence that the pollen of 

 any of the characteristic mutations differs from that of f. typica. 



It follows that mass 

 mutation in 0. pratin- 

 cola must be due to the 

 wholesale modification 

 of female gametes. The 

 relations have not been 

 worked out in the case of 

 0. Reynoldsii, which also 

 shows mass mutability. 

 There can be no 

 doubt that mass muta- 

 tion is not Mendelian 

 segregation, although 

 the two phenomena 

 have points of resem- 

 blance. Heribert- 

 Nilsson's hypothesis to 

 account for the muta- 

 bility of 0. Lamarckiana 

 depends upon the segre- 

 gation of plural factors 

 for the same character, 

 and involves such com- 

 plications as the elimina- 

 tion of all zygotes which 

 are homozygous with re- 

 gard to the presence of 

 any of the numerous 

 plural factors. Needless 

 to say, he has also relied 

 upon the doctrine of the 

 equi^'alence of male and 

 female gametes. His 

 last paper bears evidence that his faith in the equivalence of 

 gametes is beginning to waver, although he has formerly trusted so 



Fig. 14. — Mut. formosaXi. lypica (Lexington 



E-S-206XE-5-229, one of the Fi progeny): this 



cross is identical with mut. fonnosa itself (cf. 

 hg. 13); the constitution of the Fi progeny is given 



in table VI; the progenies resulting from self- 



poUination of the parent plants are recorded in 

 tables III and VI. 



