THE INHERITANCE OF FLOWER COLOUR IN PISUM 69 



Fl crosses of which four represented crosses between white and rose 

 flowered peas (Tschermak 1902). For these he (1904 a and b) pointed 

 out segregation in F-, in the ratio 9:3:4 — facts which now are well 

 known, and which afterwards have been pointed out by several workers. 

 Further examples will be given in this paper. On the presence and 

 absence hypothesis this fact is interpreted thus: the white-flow^ering pea 

 has a cryptomeric factor which alone has no visible effect, but which 

 combined with the rose factor gives purple flowers. If the rose 

 factor is missing the flowers will be white whether the other factor 

 is present or not. As far as I can find these two flower colour factors in 

 Pisum are the only ones that have been pointed out in previous litera- 

 ture. White (1917 a) thus mentions only two: A (»Salmon pink or 

 rose flower colour») and B (»purpling factor; gives with A purple 

 flowers»). Lock (1908) names the same factors C and P, Tschermak 

 (1912) in conformity with White A and B andKAJANUs (1919) R and G. 



2. MATERIAL AND METHODS. 



Already in 1898 I received from the Agricultural College of Ultuna, 

 Sweden, a pea called »Ljusröd-blommig Graärt», i. e. a light red flower- 

 ed grey pea (»grey pea» is the Swedish name for Pisum arvense), 

 a sort of pea with flowers as pale as the rose-coloured, and, accordingly, 

 with white or almost white standard, but with wings not rose-coloured 

 but light purple (See plate I). This pea will be named in he following 

 by its stock-book number, 01001. Several other lines mentioned in 

 this paper will receive similar numbers. When crossing this 01001 

 with natural whites, i. e. white forms not obtained by artificial crosses, 

 I received without exception a purple flowered Fi, just as when cros- 

 sing white and rose. But in F2 not only purple, light purple and white 

 flowers appeared, but also rose and further a fifth colour, namely 

 violet (see plate I). Phenotypically this violet can be said to stand in 

 the same relation to purple as light purple to rose. 



A closer examination of the colour inheritance in Pisum could not 

 for several reasons be started until 1911. Although the working out of 

 the different modes of inheritance at first seemed to be very simple, 

 serious difficulties were met with in the course of the investigations. 



Moreover, only limited time and work conld be given to these 

 experiments, the breeding for practical use requiring most of the time 

 and labour. 



Difficulties are often met with in the marking of the different 



