COBB AND BARTLETT: INHERITANCE IN OENOTHERA 473 



It has already been noted that some of the plants resulting 

 from the cross between strains E and C were characterized by- 

 defective leaf development, and that some defective typica indi- 

 viduals bore normal branches, which had every appearance of 

 being bud sports. In two such cases separate progenies were 

 grown from seeds of the normal and defective portions of the 

 same plant, with the very striking results summarized in table 5. 

 The progenies contain only normal plants if derived from seeds 

 borne on the normal bud sport, but if derived from seeds of the 

 defective part of the plant, they contain a considerable propor- 

 tion of plants classified as defective f. typica. The cultures are 

 large enough to afford conclusive evidence that whatever the 

 change may be that results in the production of normal from 

 defective f. typica, the change is one that may come about in 

 the somatic cells, and, once having come about, is permanent. 



As a final proof of the Mendelian nature of the segregation in- 

 dicated by the 3 : i ratio in the F2 generation, F3 progenies 

 have been grown from a large number of self -pollinated normal 

 typica individuals, in order to demonstrate the existence of 

 homozygous and heterozygous dominants in the ratio i : 2 in 

 the F2 generation. The results are not included in this paper 

 for the reason that they are still being added to and it does not 

 seem desirable to publish only a portion of the data. However, 

 it may be stated that not only do the expected classes occur in 

 • the correct ratio in the F2 generation, but also that the heterozy- 

 gotes continue to split in the 3 : i ratio. The homozygotes of 

 both classes breed true. 



Having established the non-Mendelian behavior of the charac- 

 ter-pair flatness vs. revoluteness within strain K, and having 

 shown that flatness when introduced into strain E by crossing 

 a revolute-leaved type of the latter with strain C, does act in a 

 Mendelian manner, it remains to indicate a possible explanation 

 of the phenomena, 



THE HYPOTHESIS OF HETEROGAMETISM. 



In accordance with the hypothesis of nonequivalent gametes, 

 which has been decidedly helpful in the interpretation of some 



