STUDIES ON THE GENETICS OF FLOWER-COLOURS, ETC. 
99 
hardly distinguish them exactly by their external appearance. 
Thus we have in F, 4 homo- and 7 heterozygous plants of orange colour» 
which accords, in spite of their small number, fairly well with the calculated 
numbers 3-7 and 7 - 3 respectively. 
From all above described we see that the difference between orange and 
wliite-I varieties is due to one factor, and that tins factor which we call C 
produces orange colour. 
Though I have also raised the F± generation of the above cross it will 
not be perhaps worth while to describe here its details, and it may suffice 
simply to say that it has fully confirmed the results of the former generations. 
During some generations of the above cross certain peculiar phenomena 
were often met with, which might somewhat confuse the Mendelian results 
obtained ; thus, for instance, few magenta and red progeny are produced 
from seeds taken on orange or even white plants, and also few orange ones 
from those taken on white plants, etc. (naturally seeds being taken on selfed 
flowers). These phenomena will, according to my view, chiefly, though not 
all, belong to the so-called “ reverse mutations,” and since they were observed 
in many other cases they will be pointed out each time in the course of my 
description, and discussed together later in a separate chapter (s.p. 121 ff). 
Cross II. Yelloio X white-I. (PI II, fig. 4 and 8). 
CCG G rrbb x ccggrrbb F v = CcGgrrbb. 
In order to study the genotypic constitution of the yellow variety I have 
done, firstly, its cross by the white-I (1915), and secondly, the two reciprocal 
crosses between it and the orange (1917). The /{-hybrid produced by the 
latter crosses has borne flowers where yellow and orange patches of various 
size are irregularly scattered on each petal ; in F 2 we were however impossible to 
exactly determine yellow and orange individuals, which we might have expected 
to have been produced by segregation, and the experiment was abandoned, 
at least for a time. As to the first of the crosses above cited, I was much more 
fortunate, though on account of the poor germination of seeds the behaviour of 
the /'(-generation could not be so fully investigated as might be wished for. 
The /{-hybrid made by crossing it by white-I has borne yellow flowers whose 
colour is nearly the same, or even more intense than in the yellow parent (s. 
