61 



ing to admit at least tlie close outward similarity of the genera, 

 despite their actual floristic divergency. Here are records of 

 greenhouse experiments, with actual production of hybrids 

 between //. Richardsonii var. hispidus and //. sanguinea (a 

 Mexican species), illustrated by good pictures of growing plants 

 and flowers of parents and hybrids. //. Richardsonii with 

 strongly zygomorphic, yellowish-green flowers belongs to the 

 section Ileruchea; II. sanguinea, in which the flowers are bril- 

 liant red and almost campanulate, belongs to Rhodoheuchera. 

 The hybrids resemble large-panicled //. Richardsonii, but the 

 flowers are "tubular campanulate, moderately irregular in form, 

 and pink in color with greenish sepal-tips." Segregation of F2 

 in Alendelian fashion brought out variants in height, color of 

 flowers, and character of hypanthium. These gradually sifted 

 out to about six strains, preponderantly like //. Richardsonii, 

 and it is assumed that the Sanguinea -iovms, being less hardy, 

 died out first. "Just how^ to treat, taxonomically, such inter- 

 breeding groups is something of a puzzle. Obviously they have 

 not quite attained to the full stature of species, as species occur 

 in such a genus as Carex. They have reached a similar degree of 

 morphological differentiation without attaining full biological 

 independence. They are kept distinct in nature not through any 

 biological incompatibility but merely by the accident of geo- 

 graphical isolation or, in a few cases, by differences in flowering 

 periods; and these accidental barriers, whether of space or time, 

 may at any time break down as a result of climatic changes 

 resulting in extension of range or change in the time of flowering. 

 In some instances this sort of thing appears actually to have 

 happened in the not-distant past. Thus extending ranges of //. 

 americana and of //. Richardsonii appear to have recently over- 

 lapped, and the two species are completely blended over an 

 area of more than a hundred thousand square miles, though 

 elsewhere they are perfectly distinct and are not even closely 

 related. Presumably the failure of other groups to blend in a 

 similar manner is due far more to geographical and historical 

 than to biological causes" (p. 18). 



In the accompanying map (fig. 5) these hybrids between //. 

 americana and //. Richardsonii are seen to occur in a triangle 

 extending from Michigan to Arkansas and Minnesota. Fitting 

 such plants into ordinary taxonomy is accomplished by "recog- 



