COBB AND BARTI^ETT: INHERITANCE IN OENOTHERA 475 



and /3 gametes between the two sexes. In view of the very 

 large number of modifications occurring in the a gamete, how- 

 ever, it seems a more probable hypothesis that the uncom- 

 pensated characters are borne by a group of chromosomes, which 

 do not segregate freely, but pass together to one daughter cell 

 or the other at the reduction division. Certainly such a hypothe- 

 sis is useful in the interpretation of Oenothera genetics. The 

 phenomena of matroclinic inheritance are clearly orderly enough 

 in their own way, but are just as clearly non-Mendelian. 

 As Shull (see footnote ii) has said: "Fundamental difficulties 

 are encountered whenever attempts are made to apply to the 

 Oenotheras rules of genetic behavior which are readily demon- 

 strated in other groups of organisms. Equal confusion has 

 arisen by the application of genetic experiences with the Oeno- 

 theras to species in which typical Mendelian phenomena appear. 

 A hereditary mechanism must exist in Oenothera fundamentally 

 different from that which distributes the Mendelian unit-charac- 

 ters." 



According to the hypothesis of nonequivalent gametes, the 

 Oenotheras should show Mendelian inheritance in cases where 

 the factors concerned are carried by both a and /3 gametes. In 

 case the factors are carried by only one kind of gamete, non- 

 Mendelian inheritance must be the rule. The a and (3 gametes 

 may be conceived of as due to the distribution to the daughter 

 cells at meiosis of a distinctly maternal or distinctly paternal 

 set of chromosomes. In other words, instead of a free segre- 

 gation of chromosomes taking place, one must assume that the 

 haploid set of chromosom^es is reconstituted, as far as the fac- 

 tors peculiar to a and (3 gametes are concerned, just as it entered 

 into the zygote at fertilization. As for the Mendelian charac- 

 ters, we assume that they are represented by factors carried in a 

 residue of freely segregating chromosomes. 



The conception of heterogametism may be reduced to the 

 following series of propositions: 



(i) In heterogametic species the zygote normally results from 

 the conjugation of unlike gametes (a and /S). 



