168 
potency". The „inhibiting factor"^ if such there be, need not 
even be genotypic in nature^ but may be the result of some 
somatic quality of the parent (such for instance as its State of 
health) projected to the offspring through influences surrounding 
the latter during their embryonic development. 
Summary. 
Grosses between Bursa (Capsella) hursa-pastoris simplex and 
B. Heegeri Jieteris have demonstrated the existence of two genes, 
Ä and B, which determine the differentiating characteristics of 
the rosettes and which result in the production of four forms in 
the F2 in the ratio 9:3:3:1. 
They seem to indicate also the presence of two genes, C 
and J), each of which is independently responsible for the bursa- 
pastoris-tjipe of capsule. The Heegeri-tjipe appears only in the 
absence of both C and D. On this basis the two forms should 
appear in the Fg in the ratio 15 : 1. The observed ratio was 
21.9 : 1. In the F3, some famihes should give ratios of 3 : 1, and 
other families should again give ratios of 15 : 1. The observed 
ratios in Fg were 4.67 : 1, and 22.2 : 1. These observed ratios 
show a corresponding deviation from the theoretical ratios, and 
are thought to indicate the action of some modifying influence, 
such as „selective elimination", distorting the results of an other- 
wise normal Mendelian segregation. 
A defective ratio in the rosettes of one family is shown to 
be due to failure of dominance, and certain facts are presented, 
which indicate that this failure of dominance is attributable to a 
less „relative potency" of the gene Ä which determines the 
tenuis- chsireLcter. This change in the relative potency of Ä may 
be due to a decline in the efficiency of the gene itself, or to 
the Operation of some other factor or condition which offers an 
increased resistance to the development of the few^^/s-characters. 
