140 Proceedings of the Biological Society of Washington. 



tacitly doubted the correctness of this conclusion, but there has 

 been no critical study of the group as a whole. In particular, 

 the common plant of southern California, passing for the most 

 part as C. myriophylla, has long been known to depart widely 

 from current descriptions of this species. More recently this 

 discrepancy has again been brought to the writer's attention by 

 Mr. George L. Moxley, and by several other California botanists 

 quite independently, and the present paper has resulted from 

 an effort to clear up the existing confusion in the whole series 

 of related or misidentified forms occurring in the southwestern 

 states. The conclusions reached are altogether at variance from 

 Eaton's judgment above quoted, but are in the main satisfac- 

 tory and appear to be fully substantiated by the great amount 

 of herbarium material now available. 



The facts appear to be, briefly, that C. myriophylla, an 

 andine species of North and South America, does not extend 

 north of Mexico and must be excluded from the United States 

 flora; that C. villosa Davenp., of the Mexican Border region, is 

 a nearly related but well defined species; that the plant of 

 southern California and adjacent territory represents a variable 

 undescribed species (C. Covillei), including a northern, mainly 

 coastal form which further study and material may show to be 

 specifically distinct; and that the southwestern plants referred 

 mostly to C. Fendleri actually pertain to two species, C. Fendleri 

 and an undescribed species (C. Wootoni), similar in general habit 

 but easily recognizable upon characters of the scaly covering 

 of the under surface. 



The distinguishing characters of these species and of C. Cleve- 

 landii, which has sometimes been confused with C. Covillei, are 

 given in the accompanying descriptions and key. For the pur- 

 poses of the present paper it is not necessary to deal critically 

 with C. myriophylla, which, excluding the United States plants 

 heretofore so referred, is taken in its traditional sense. Thus 

 restricted, C. myriophylla is still a variable species, possibly com- 

 prising several forms worthy of separation . The tropical material 

 at hand is ample and will serve as the basis of a later paper. 

 Its segregation does not affect the status of species occurring 

 within the United States. 



