The evidence in support of these hypothetical formulas is derived in 

 two ways from a germinal analysis — by means of (i) selling, and (2) 

 hybridization. The evidence is considered separately under these two 

 headings. . . -.^i 



BEHAVIOR OF COLOR TYPES WHEN SELF-FERTILIZED 



Type 1 White (Plate II) 



Description of the type. — The white type is a pure white and a true 

 albino. The star-shaped rays in the corolla may in some cases be slightly 

 yellowish, but so pale as to be hardly different from the ground color. In 

 other cases they may be flushed with pink or may have a definite spot 

 of lavender. ■, 



The albino type can be accounted for hypothetically by the absence 

 of sufficient chromogen to fonn any anthocyanic pigments (cc or Cc) or 

 by the lack of sufficient oxidase (R) to react with C and form anthocyan, 

 without which combination the other genes cannot act even tho they may 

 be present. Thus any fonnula lacking either C or R, or even one having 

 C in a simplex condition, might represent a white. 



Theoretical consideration oj data. — Thirty-three white plants were selfed 

 and yielded an aggregate progeny of 262 whites. No colored types 

 appeared. Chemical tests on soine of the white flowers seemed to indi- 

 cate the presence of both jieroxidc and peroxidase in the tissues of the 

 corolla, bvit a lack of chromogen. Certain it is that these plants lacked 

 some essential part of the mechanical device for the production of pigment 

 — either C or R, or both. The presence or absence of other genes could 

 be determined only by cross-breeding. 



Other whites threw colors when selfed as shown in table 2 ; 



128 



