VULGARIS L. AND IN PH. MULTIFLORUS WILLD. 125. 
Conclusions regarding the genotypical constitutions, that are 
realized in these phaenotypical groups, by which the segregation 
may be explained, are not at all easy because of the relatively little 
number of individuals. From the previous crosses however some 
suppositions have been derived, that may be justified here too: 
the Kievits-race would possess a factor K responsable for the red- 
violetmarbling, while in it the factor L for grey-brown is absent; 
this Kievitsfactor does not occur in the Rotjes-race, but the L-factor 
is one of its characteristics. The segregation however shows to be 
more complicated than a dihybrid, for the K- and L-factors only 
are not sufficient for explaining the occurrence of phenotype II 
(blue-violet colouring). Probably the Rotjes-race is in possession 
of still another factor: Bl, a factor that is changing the reddish 
violet into blue, and that remains invisible in the kotjes-race. This 
Bl-factor would be visible only, if reddish-violet is present and thus. 
it cannot make its appearance in types IV and V, remains therein 
in a cryptomeric state. Now, if types with and those without L are 
grouped, those with and those without Bl, those with and those 
without K, one would expect to find a 3:1— segregation in each 
of these cases. The results dont agree with this expectation: 


n Sie hour ew th El wiihoue Bl it Ke.) without K 
Wee aad IV IL dlvend Vil and II | til) | Tl andl IV and-V 
| | | 
DR Se er eee 120 ins a 1014 
These numerical proportions may be considered as rather sufficient 
with regard to the factors Bl and K (expectation 40:13, resp. 
50:17); those of the L-factor however show a considerable deviation 
from the expected 3: l-proportion, so that they lead to the conclusion, 
the L-factor must be in a cryptomeric state in some individuals. 
Probably this will be found to be the case in the bl-individuals, 
where thus the Bl-factor is missing, for the utterance of the L- 
factor cannot be dependent upon the presence of the K-factor, 
because of its phaenotypical perceptibility in the Rotjes-race, where 
the K-factor is entirely absent. So the case will find a rather good 
explanation, if one supposes the Bl-factor to be cryptomeric in those 
beans, that are missing the K-factor and the L-factor as cryptomeric 
when the Bl-factor is absent. Type I will be represented than by the 
