FLOWER AND SEED COLOUR IN LUPINUS 315 



F, repeats this di-hybrid segregation with: 
1160 blue : 586 violet : 605 tinged blue 


expect: 1175,50 587,15 587,75 
15,50 | TES Reese ec 
D/M fe 2] = U,08 2] = (77 
There are additional deviations from normal segregation to note. 
Constant blue flowered families, as well as 3:1-segregating families 
from blue parent plants, are wholly absent. All families from such 
parent plants give the di-hybrid type of segregation in the ratios 1:2:1. 
All families from violet and tinged blue parent plants are, furthermore, 
constant and do not show any segregation. All these deviations in 
the proportions of the families and of the phenotypes are charac- 
teristic of an absolute or close linkage. 
It is dilution and the violet colour which, phenotypically seen, 
never or rarely enter into combination. Genotypically it is explained 
by assuming a repulsion between the factors B and F; the pheno- 
typical correlation between violet and fully developed colour should, 
in other words, depend on the closeness between the factors represen- 
ting bluish red and full colour according to the chromosome-theory. 
This depends of course on the construction of the factorial scheme. . 
The violet colour caused by the factor V appears only when the B- 
factor has been excluded, and vice versa. Consequently, this inverse 
proportionality between the phenotypical manifestation and the ge- 
notypical mode of characterization will be found in all the following 
crosses, where linkage occurs. This would have been avoided had 
the CastLe-MorcGan method of factor denotation been adopted (cp. 
Morean 1915, pag. 253). The genetical formulae would have been 
constructed in the following way: Blue = wild type; bluish red is 
caused by mutation of the B-factor, denoted b; we denote further violet 
colour by v instead of bV, and pure red by bv und not by bvR; 
tinged colour is caused by the absence of F (=f) as before. The re- 
pulsion in cross 8 would then be caused by the close linkage between 
the factors f and v. The fact that our case corresponds in colours 
and in the mode of segregation with analyses previously made favoured, 
however, the adoption of the traditional factor denotation. 
The degree of linkage can not be determined, as no recessives 
appear. It may be absolute, but partial linkage of a relatively high 
degree is also conceivable. The results obtained allow but an approxi- 
mate estimate of the minimal value of its strength. The total number 
of individuals of the di-hybrid segregation in F, and F, is 3317. 
Hereditas II, 21 
