88 



THE AMERICAN NATURALIST [VOL. XLIV 



If, on the other hand, the assumption be made that, in 

 Lychnis, absence of B is dominant over its presence, the 

 relative positions of R and B may remain the same as in 

 Lathyrus, Matthiola, etc. For in this case R and C will 

 be the two factors necessary to the production of red 

 anthocyan and B the "bluing factor" which is added to 

 it to form bluish anthocyan, the difference between 

 Lychnis and Lathy rus being simply that the B which 

 may be looked upon perhaps as a factor for alkalinity or 

 for an oxidizer, is too weak in its activity in Lychnis to 

 produce its characteristic effect except when present in 

 double quantity or strength, as it is when in the homozy- 

 gous state. 



The data now at hand do not make possible a decision 

 as to whether presence, or absence of the color-modifier, 

 is dominant in Lychnis, as each of these assumptions may 

 be shown to fit all the facts involved in Table I. The 

 gametic formulae of the six plants involved in these eight 

 crosses are presented in Table III., to enable a compari- 

 son of the two methods of explanation. 



TABLE III 



The fact now demonstrated that the purple color in 

 Lychnis is due to the presence of at least three distinct 

 genes instead of only one, as originally assumed by me, 

 has served to elucidate several difficulties which had 

 been encountered. It may be recalled that a rather 

 large range of supposedly fluctuating variability in the 



