118 SIRKS, THE COLOURFACTORS OF THE SEEDCOAT IN PHASEOLUS 
given above; Type I had a more or less intense reddish violet 
marbling on a yellowbrown groundcolour as had the F,-individuals ; 
according to the prevailance of the violet or of the brown colour, 
this type has been divided further into five subtypes (I a—e); 
subtype Ic agrees with the F,. Type II was selfcoloured yellow- 
brown; in some cases however the seedcoat had a violet glow 
upon the brown colour, so that two subtypes could be discerned: 
Ila yellowish-brown, and IIb the same with a violet glow. Type 
Illa was the same as the well-known Wagenaar-race, while IIIb 
had the same seedcoat but with a violet glow. The plants, classified 
as type IV had the same colours and the same pattern as the 
Kievits-race; as in type I here also the intensity of violet showed 
great differences between the various individuals, belonging to this 
group, so that a division in five subtypes was applied: pl. II fig. 
19 shows a bean of the lightest of these subtypes, IVe. Types V, 
VI, VII and VIII were wholly without violet and without yellow- 
brown in their phenotypes: V the lemoncoloured like the Citroen- 
race, VI yellow and white marbled (pl. II fig. 20), VII pure white 
and VIII a greyish white seedcoat, but possessing a brown navelring 
(pl. I fig. 10, probably identical with Kooimans H-type). The indi- 
duals of the seven Fo-families, that were grown, were distributed 
over these various types in following frequencies: 
(Cf table page 119) 
At first sight these numbers may seem to be of little promise; 
if however the numbers of groups I, II, III a.s.o. are taken one 
for one, neglecting the further divisions Ia, Ib, Ic, a.s.o., arather 
good numerical proportion can be found therein. There are types . 
with and types without the brown navelring (factor D), types with 
and types without the yellowbrown colour of the whole seedcoat 
(factor G), types with and types without the Kievits-pattern (violet 
marbling, that- may provisionnally be ascribed to a factor K) and 
finally among the recessive lemoncoloured beans uniformly coloured 
and marbled individuals. This lastmentioned segregation may be 
explained rather sufficiently by the supposition of KOoIMAN, that 
a chromogenic factor B in homozygotic state causes an equal 
colouring, in heterozygotic presence however the inconstant type. 
of marbling. It is however unclear, why this difference between 
