256 



KARL B. KRFSTOFFERSON 



TABLE 1. The cliantctcrs o/ jlowcr constnicfion in the parent lines 

 and in the Fi-generation. 



The Fa-generation of the cross 2 X 10 showed segregation in 

 herkogamous and autogamous forms. In order to determine the ratios 

 isolations with parchment hags were made; 68 individuals did not 

 develop any seed and 28 were autogamous. The variation seemed to 

 he quite discontinuai. All autogamous plants had the capsules well 

 developed with a rather large numher of seeds. In the herkogamous 

 plants no sign of development of seed was to he seen. 



The ratio hetween herkogamy and autogamy is rather close to 

 the ratio 3:1, viz. 2,s()4 : l,i96. The theoretical ratio is 3: 1 + 0,1758; 

 the number of autogamous plants is thus seen to be somewhat high. 



The deviation from the 3 : 1 ratio may be explained by the 

 disturbing influence of the presence of thrips in the flowers. They 

 have sometimes been observed in the isolated flowers. It is possible 

 that they have entered through small holes in the parchment bags, or 

 there may be eggs on the buds of the flowers. Such isolations have 

 of course been rejected. I have also repeated the isolation if only 

 the first developed flower was found to develop seeds. In such 

 cases the plant was often found to be herkogamous. 



Thus, the deviation from the 3 : 1 ratio may easily be explained 

 by assuming a self-pollination by small insects; the monohybrid 

 segregation as such is more difficult to understand when the great 

 variation of the floral parts is remembered. The herkogamy may be 

 due to three characters, viz. the direction of the entrance to the 

 stigmatic chamber, the development of the pollen-magazine and 

 labellum. 



These three characters showed a continuous variation in Fo. As 

 to the first mentioned it seemed to be a transgressive one. They 



