KNOWLEDGE OF HEREDITY IN PISUM. 103 



(iv.) Tlie behaviour of the two pairs of allelomorphs C-c 

 and P-p in combination is most readily observed by noting 

 the colour of the flowers of the different plants, though it was 

 also possible in every case to recognize the class to which a 

 plant belonged on examining either the axils or the testas. 

 The two pairs segregate quite independently, P being totally 

 invisible in the absence of C, as the following evidence shows. 



The varieties uned were : — 



(1) Purple — Purple Suga.r Pea ; Purple Podded Pea ; 



Maple ; and a Purple Field Pea. 



(2) Red — Irish Mummy Pea. 



(3) White — Laxton's Alpha ; Veitch's Perfection ; Sun-' 



rise ; British Queen ; Victoria Marrow ; Tres nain 

 de Bretagne ; and various extracted whites. 



In the case of crosses between these original strains, purple 

 crossed with white always gave all purple in F 1, and in F 2 

 yielded a proportion of 3 purple plants to 1 white. 



Purple crossed with red gave all purple in F 1 , and in F 2 

 3 purple to 1 red. 



Red crossed with white invariably gave all purple in F 1 , 

 and in F 2 gave 9 purple : 3 red : 4 white — in a particular 

 case 141 purple : 43 red : 65 white.* 



Thus, among extracted whites derived from the cross between the 

 purple sugar pea and different white strains I found no example which 

 did not yield exclusively piu-ple spotted offsprmg when crossed with 

 gray. The experiment, howevfer, should be repeated, since the whites 

 used were F 3 and F 4 and not F 2 ; also the development of the purple- 

 spotted character varied considerably in coloured F 2 plants. 



Bateson found differences between the product of reciprocal crosses 

 in which tliis character was concerned. Such differences can be accounted 

 for on the supposition of segregation of the S-s pair in the white strain 



used. 



The account given in the text iindoubtedly holds good in certain 

 eases, and the possible exceptions are not yet fully established as such. 



* That is to say, all the white strains examined contained homozygous 

 P in the luasked condition. The absence of wliite varieties which do 

 not contain the factor for piirple may be to some extent understood 

 when we remember the necessary invisibility of this factor in wlute 



plants. 



In the course of the history of modem cultivated peas, white-flowered 

 plants, which must have arisen in the first instance from pm-ple- flowered 

 parents by the loss of the factor C, could be, and doubtless were, selected. 



