9 



that the black, brown and yellow varieties 

 should all contain the same latent characters, 

 and I am forced to the conclusion just stated, 

 that the novelties which appear in these hy- 

 brids are directly derived from the white 

 parent. The new characters thus lose the 

 mystery that otherwise invests them. They 

 appear not as released prisoners or awakened 

 sleepers, which had for some inconceivable 

 reason remained inactive, but are combination 

 phenomena. 



The color characters of these beans are not 

 dependent upon a single pair of units, but 

 upon three pairs, viz. : 



1. Pigment vs. no pigment. 



2. Modifier which changes pigment to 

 purple vs. no modifier. 



3. Mottled color pattern vs. self color. 



Of these three characters, the brown and 

 yellow beans contain only the dominant pig- 

 ment character and might be represented by 

 the formula, P~bm; the black bean contains 

 the pigment and the modification to purple, 

 PBm; and the white bean contains both the 

 modifier and the mottled color-pattern but no 

 pigment, pBM . It is only because of the lack 

 of pigment that these two characters possessed 

 by the white bean are not apparent. They 

 are latent only in the sense that they are 

 invisible. Whenever the ' White flageolet ' 

 is crossed with any variety of self-colored 

 bean, the three dominant allelomorphs, PBM, 

 are brought together with the result seen in 

 these hybrids, namely, a first generation char- 

 acterized by dark purple mottled seeds. 



The great advantage of this explanation 

 over that of Tschermak is that it brings these 

 apparently aberrant results into harmony with 

 typical Mendelian cases, and allows with a 

 reasonable degree of accuracy, a prediction as 

 to the composition of subsequent generations. 



On the assumption that the F 1 hybrids be- 

 tween the ' White flageolet ' and either of the 



