454 Genetics of Sex Determination 



place, the sex-determining and the monoecism-controlHng mechanism, 

 the latter of which we described as a patterning of primordia within 

 a flower bud. 



The Streptocarpus crosses ( Gesneriaceae ) of Oehlkers (1938- 

 1952) represent the best analyzed case. The species are monoecious, 

 with two functioning and three rudimentary anthers ( staminodia ) . 

 Of the six species investigated, four species behave alike and the 

 other two form a separate group. As examples, the crosses between 

 Rexii and Wendlandii are presented (W = Wendlandii; R = Rexii; 

 2 always given first place). 



Fi R X W dioecious, like parents 



Fi W X R anthers reduced to staminodia 



RF2 (W X R) X R half like Fi, half normal gynoecia, but, instead of anthers, 



all transitions from anthers to carpels 

 RF2 (R X W) X W half normal monoecious, half externally the same, but with 



rudimentary ovaries 



To this group were added later crosses with another monoecious 

 species, polyanthus (P). The results are as follows: 



Fi P X W like parents 



Fi W X P like parents 



F2 (W X P)2 3 like Fi : 1 without anthers, i.e., like Fi W X R 



RF2 (W X P) X P 1 like Fi : 1 without anthers 



F2 (P X W)2 all like Fi 



The following formal explanations have been proposed for the 

 first group of facts ( the second group to be considered later ) . Oehlkers 

 follows Correns in assuming that the genome contains a general AG 

 complex of female and male sex determiners. In Rexii the G factors 

 are superior to the A factors and vice versa in Wendlandii. The 

 plasmon in either case must pull in the opposite direction, because 

 both forms are monoecious. Thus R plasmon pulls in the male direc- 

 tion; W plasmon, in the female direction. In the backcross (W X R) 

 X R, the W plasmon, pulling in the female direction, becomes com- 

 bined with genomes working in the same direction; therefore, female 

 intersexuality occurs in half of the offspring. This requires, of course, 

 that the assumed AG complexes segregate like single loci. If we try to 

 visualize this in detail, insurmountable difficulties arise. However, 

 discarding this unworkable AG hypothesis, we may postulate a single 

 pair of Mendelizing alleles and a modifier, present in the R genome, 

 which pulls, if homozygous, in the female direction, but only in a 

 proper plasmon, which is found in W. One could mention in favor of 

 such a modifier that Oehlkers found in the same crosses a modifier for 



