i6 Bulletin 392 



type of some one or more genes essential to the formation of anthocyanic 

 pigment. If the full pink is supposed to be due to a double portion of the 

 same gene (that is, a duplex condition) that causes the pink flush of the 

 tinged white type, the pink type may be formulated as CCRR. This type, 

 being homozygous, could then give only pinks when selfed. 



Fourteen pink plants were selfed and in a progeny of 127 plants only 

 pinks were found. It is a true-breeding type, as the hypothesis stated 

 above demands. 



Plant 211-1 was recorded as pink and checked as such. When selfed 

 it threw 3 magentas, 3 mauves, 8 pinks, i tinged white, and 12 whites. 

 It must have been a mauve (genotypically) with the formula CCRrXxII, 

 which could throw all these types when selfed. Mauve could very easily 

 have been mistaken for pink and recorded as such. 



Type 4 Mauve (Plate I) 



Description of the type. — The type called mauve is slightly darker and 

 bluer than that called pink. The two are often so nearly alike that it is 

 difficult to be certain to which type a flower belongs. This may, no doubt, 

 have given rise to some experimental errors in recording the type, as in 

 the case of plant 211-1, discussed above. 



Theoretical consideration oj data. — The data showing the behavior of 

 mauves when selfed is very meager. Five plants were selfed and yielded 

 an aggregate progeny of 80 plants, of which 61 were mauve and 19 were 

 pink. The largest pedigree contained 59 plants, of which 42 were mauve 

 and 17 were pink. Other plants threw only pinks, but the numbers were 

 very small. The significant fact here, however, is that no types other 

 than mauve and pink were thrown. 



The mauve type may be considered as the pink genotype plus an added 

 gene X which reacts with R to give a deeper and bluer hue. If a mauve 

 plant were of the genotypic constitution CCRRXX, it should breed true; 

 if simplex for X, it should throw three-fourths mauves and one-fourth 

 pinks but no other types. This seems a plausible explanation of the data 

 given above. Some of these plants may have been either simplex or 

 duplex for X, since the numbers of their progeny were too small to demon- 

 strate which was the case. 



Type 5 Magenta (Plate I) 



Description of the type. — The type called magenta is ver\' similar to 

 the mauve type, but tlie color is deeper, richer, and more intense. It is 

 a brilliant color and does not change when the flower withers. It occurs 

 also in a striped or flaked pattern on a white background, as discussed 

 under a later heading (page 27). 



