FLOWER AND SEED COLOUR IN LUPINUS 327 

and Graminae, which are at the same time among the best studied 
with regard to genetics. Of course, the method has less value in many 
cases, as for instance in coupling between homomeric factors, in case 
of poor vitality on the part of recessives, in the event of a marked 
decrease in the germination power of the seeds when stored for 
sowings in successive years. It is capable of doing good service, 
however, when discriminatively used. 
The result of the observations as to the seed colour in cross 9 
remains to be stated. All isolated plants in F, and F; have been 
examined with regard to this point, and the colour of the flower has 
been noted at the same time. All blue and bluish red plants, amoun- 
ting to 179 and 67 respectively, had earth-brown seeds, and all violet 
and pure red plants, amounting to 68 and 2 respectively, had rust- 
brown seeds. This agrees with the assumption that the rust-brown 
seed colour depends on the same factor as the pure red flower colour 
(factor R), and that the addition of the factor for violet colour (V) 
does not influence this seed colour, which remains rust-brown even 
when the flower colour passes into violet. It also agrees with the 
assumption that the earth-brown seed colour is caused by the B- 
factor when the fundamental factor R is present at the same time, 
while it is quite independent of the presence or absence of the V- 
factor. The assumed correlation between the seed colour and the 
flower colour is further strengthened by the fact that all families, con- 
stant with regard to pure red flowers in F3 and F4, also showed con- 
stancy with regard to rust-brown seed colour. The constant violets 
and the families segregating in violets and pure reds were also constant 
as to the rust-brown seed colour. Only a small number of families, 
however, were examined in this respect. In addition numerous cases 
have been investigated although not listed, and the correlation in 
question has been found to hold true in every case. 
Cross 10 (tables 25—26). 
Bluish red (RRBBvvFF) X tinged blue flower colour (RRBBVVff): 
no difference in seed colour. 
F, was slightly diluted blue. 
F, gave the following result: 
77 blue : 32 ting. blue : 41 bluish red:6 ting. red 
expected | 
acc. 19%3:3:1 87,15 29,25 29,25 9,75 
D/Mı War 275/54 =—0 5 1175/54 = 2,18 del, 
