1907.] Floiver Colour in Antirrhinum majus. 295 



three factors only ; the introduction of a fourth factor would give magenta 

 and ivory individuals. 



A similar diagram constructed for four factors shows us that, as regards the 

 composition of the F 3 they give on selfing, there are 16 kinds of magentas 

 in F 2 . Among 13 magenta plants taken at random in F 2 and selfed, 8 out of 

 the 16 kinds were found. (For Case 1, Table II gives F 3 resulting from all 

 the F 2 individuals selfed.) The remaining 8 would doubtless occur among a 

 larger selection of F 2 individuals. 



Of magenta delilas, there are theoretically 8 kinds; 6 plants taken at 

 random gave 2. 



Of crimsons, theoretically 8 ; 5 plants gave 3. 



Of crimson delilas, 4 ; 3 plants gave 2. 



Of ivories, 4 ; 6 plants gave 3. 



And of yellows, 2 ; 3 plants gave both. 



A corresponding diagram can easily be constructed for Case 2. 



From such a diagram one learns that the F 2 magentas are theoretically of 

 8 kinds; 11 plants taken at random gave 7 out of the 8. (Table III gives 

 F 3 for Case 2.) 



Of magenta delilas, 4 ; 4 plants gave 4. 



Of crimsons, 4 ; 6 plants gave 4. 



Of crimson delilas, 2 ; only 1 plant was taken. 



Two white plants gave whites only. 



Similarly, from a diagram constructed for Case 3, we find there are 8 possible 

 magenta delilas ; 5 plants gave 5. (Table IV gives F 3 for Case 3.) 



Of crimson delilas, 4 ; 3 plants gave 2. 

 Of ivories, 4 ; 1 plant only was taken. 

 Of yellows, 2 ; 1 plant only was taken. 



Finally, from a diagram for Case 4, we find there are 4 possible magenta 

 delilas ; 6 plants gave 2. (Table V gives F 3 for Case 4.) One white plant 

 gave whites only. No other F 2 plants were taken. 



It is clear also from Diagram I that there are 8 possible albinos. Various 

 yellows this year have been crossed with a number of F 2 whites as a means of 

 testing for the characters which may be borne by an albino. 



White x Crimson Delila : — 



Both these types were found to be breeding true. 

 Fi consisted of magenta delilas only. (Table I.) 



