STUDIES ON THE GENETICS OF FLOWER-COLOURS, ETC. 
103 
Cross IV. White-I X white-1 I and vice versa, etc. (PI II, fig. 8 and 7). 
ccrrbb x cell BUB F x = ccRvBb. 
At tlie beginning of my breeding experiments I have conducted the crosses 
between the various white varieties, because it was thought not to be impossible 
that the cross of two certain whites might produce the progeny with coloured 
flowers, as in the classical example of Sweet Pea. For instance, the cross 
between white-I and white-II and its reciprocal have been done in 1915 ; 
all Vj-oiïspring were found to bear white flowers in both cases, and it was 
the same in I^-progeny. The crosses between all white varieties, including 
white-III also, made in various ways agree in the fact that they never give 
rise to coloured plants, both in F x as well as F 2 , and evidently in any 
further generation. The treatment of flowers of all these varieties by ammonia 
vapour according to Miss Wheldale 1 and Shibata’ does not give yellow 
reaction, indicating that they do not contain flavones. All white plants are 
cc, because the presence of C produces the orange colour, as above stated. 
Cross V. White-I x magenta, and vice versa. (PI II, fig. 8 and 1). 
ccrrbb x CCRRBB F x = CcRrBb 
The cross, white-I x magenta and its reciprocal were made in 1916 and 
1917 respectively. In both F r liybrids we see that magenta is almost perfectly 
dominant to white (s. the Table of Colours, p. 96). On account of very 
poor germination of seeds the number of individuals is rather small, especial¬ 
ly in F 5 , but the actual results agree fairly well with the theoretical expectation. 
The Table IV indicates the results of the F 2 - and F -generation, (s. p. 104). 
We have thus in F., homozygous magentas (CCRRBB): magentas 
segregating into 3 magentas and 1 orange ( CCRrBb ): magentas segregating 
into 3 magentas and 1 white ( CcBJlBB ): magentas segregating into 9 
magentas, 3 oranges, and 4 whites ( CcBrBb ) in the ratio 1: 1 : 1 : 4, their 
expected numbers being 0‘8 : P6 : 1*6 : 3T respectively. Furthermore, we 
have 5 homo- and 6 heterozygous oranges, while their theoretical numbers 
are 3 - 65 and 7'30 respectively. Of 7 whites 5 have produced only whites. 
1 Journ. of Genetics, Vol. 4, 1915, p. 113. 
2 Got. Mag., Tokyo, VoL 29, 1915, pp. 121-122. 
