[Reprinted from THE AMERICAN NATURALIST, Vol. XLIV., Feb., 1910. | 



COLOE INHERITANCE IN LYCHNIS DIOICA L. 1 



DR. GEORGE HARRISON SHULL 



Two years ago I showed that in Lychnis dioica L. the 

 purple-flowered form behaves in normal Mendelian man- 

 ner when crossed with the same type or with the typically 

 white-flowered form of the same species (Shiill, 1908). 

 In subsequent work it has been discovered that the 

 purple-flowered plants do not form a single unit-group, 

 but that there are at least two distinct types, one of 

 which has more bluish-purple flowers, the other more 

 reddish-purple. No notice had been taken of such varia- 

 tion in the color characters until last year, although it 

 had been observed that there was some variation in the 

 intensity of color in different plants, and these had been, 

 to a slight extent, recorded in terms of intensity, e. g., 

 as " light, " " medium " and "dark." Last year several 

 individuals were observed so noticeably distinct because 

 of the bluish character of their flowers, that an effort was 

 made to determine the relationship of this light bluish- 

 purple color to the more common reddish-purple, and 

 several crosses were made representing the combination 

 of "blue" and "red," using a single red-flowered indi- 



TABLE I 



Read before the Botanical Society of America at Boston, December. 



1909. 



83 



