COLOK CHAPiACTEE. ' 3 1 



flowered species of larkspur. In respect to color, tliey present 

 a strong contrast to tlie more numerous deep blue species. 

 In the books they are made to form a group by tliemselves, 

 and the group is characterized by tlie color of the flowers, 

 that being the most distinctive mark which the several species 

 collectively exhibit. The grouping is natural. These red- 

 flowered kinds are in other respects more like one another 

 than any one of them is like any of the blue- flowered species ; 

 but the color is received as being the most significant charac- 

 teristic of the group as such. No one is likely to dispute that 

 Lobelia cardmalis and L, splendens are more closely related 

 to each other than either of them is to any blue-flowered 

 species ; and here also it is the color of the corolla which the 

 phytographer seizes upon as chiefly indicative of their close 

 consanguinity. It at once binds them together as a group, 

 and in so far separates them from their blue-flowered kindred 

 of the same genus, In Gerardia, and in a hundred other like 

 instances, there is a purple-flowered natural series of species, 

 and a yellow-flowered series ; and the color is received as a 

 subgeneric character. 



In that natural assemblage of genera known as the Asteroid 

 tribe of CompositiB the color of flowers is accepted, if not 

 without some exceptions, generally at least, in the full dignity 

 of a generic character. Asa Gray made even a higher 

 use of it. In the Synoptical Tlora of North America he 

 distributed the thirty-five genera of this tribe under two 

 subtribes which he called Homochromese and Heterochrc- 

 me?e ; the former including a series in which the flowers 

 of both disk and ray are yellow, the latter comprising those 

 ivA^in the disk-flowers are yellow, while those of the ray are 



of some other color, such as white, red, purple or blue. 

 In the light of prolonged field experience in tbe study of 

 American Composit.'e— an experience running through many 

 years, and covering a greater and more varied extent of terri- 

 tory than even Nuttall traversed— I perceive, unmistakably I 

 think, that in the treatment of the AsteretB to which I now 

 refer, there is laid upon the color character greater weight 



