STUDIES ON THE GENETICS OF FLOWER-COLOURS, ETC. 
101 
the segregation into 9 oranges and 2 whites, which agree fairly well with the 
calculated numbers, 8’25±1*43 and 2 , 75±1 , 43 respectively, and may be 
represented by Ccggrrbb , whilst another orange is homozygous and has the 
constitution CCggrrbb . — As above stated, of three yellows of which I could 
have examined the offspring one has thrown only 1 yellow, and is quite 
useless for our experiment. The second has segregated into 6 oranges and 1 
white, and since there will be no yellow which will show such segregation 
without throwing any yellow at all, it seems to me to be probable that the ratio 
is here really 9 yellows : 3 oranges : 4 whites, of which no yellow did germinate 
(i.e. CcGgrrbb). The third has produced 5 yellows and 1 white: it has 
segregated either in the ratio of 3 yellows and 1 white (i. e. CcGGrrbb), 
or in that of 9 yellows, 3 oranges, and 4 whites like the second, of which 
no orange did germinate. No homozygous yellow ( CCGGrrbb ) came under 
my observation, and this is not to be astonished, because we should have 
only 1 such out of 9 yellows. Our conclusion is therefore that we have in 
our case in F, the 9:3:4 type of segregation, that the yellow variety has in 
respect to its flower-colour one more factor than the orange, and that con¬ 
sequently the cross between yellow and white-I varieties is based on the two 
factors difference, viz. C and G. 
Cross III. Red X orange and vice versa (PI II, fig. 2 and 3). 
CCJUibb x CCrrbb F L = CCRrbb 
The ifl-hybrid has red flowers whose colour intensity is almost similar to 
that of the red parent (s. the Table of Colours, p. 96). The i^-and ./'(-offspring 
are composed as in Table III (s. p. 102) :—• 
As will be seen from this Table, though 6 out of 8 oranges have 
given rise in F ?J exclusively to oranges ( = 70), one of the remaining two has 
thrown 3 oranges and 1 red, and another 5 oranges and 20 reds. Despite all 
such facts these two orange parents (i. e. F, plants) are considered to be 
homozygous like all others, so that all 8 oranges in F, are to be expressed by 
the formula CCrrbb (s. Table VIII, Nos. 19 and 11, and also discussions about 
them in Chapter “ Mutations, etc.,” IV and II respectively, p. 128 and p. 127). 
As regards 11 reds in F, whose behaviour in F$ I could have examined 
6 were proven to be homo-, and 5 to be heterozygous (expected, 365 
