Baker — A Revision of the Elephantopeae. — /. 49 



They may be separated as follows: — 



A. stems leafy, lower cauliae leaves like radical ; paleae of pappus long 

 triangular, gradually narrowed into the setae; vestiture of compara- 

 tively few loosely spreading hairs. Carolinianus. 

 AA. Stems scapiform, with a rosette of leaves at the base; cauline leaves 

 differing widely from radical. 

 B. Paleae of pappus long triangular, narrowing gradually into setae; 

 plant distinctly pubescent. 

 C. Pappus much longer than achene ; glomeruli pubescent, but not 

 densely canescent villose; heads long (10-12 mm.) with inner bracts 

 of involucre acuminate ; stems usually little branched above. 



tomentosus . 

 CC. Pappus usually about as long or shorter than achene; glomeruli 

 densely canescent villose in typical forms ; heads shorter (6-8 mm.) 

 with inner bracts of involucre obtuse submucronate ; stem usually 

 much branched above. elatus. 



BB. Paleae of pappus short and broad, suddenly narrowed into the 

 setae; plant nearly naked. nudatus. 



1. Elephantopds Carolinianus Willd. 



Sp. PI. 3 : 2390. 1804. 



This is the commonest and most easily recognized species 

 in North America. The variety violaceus (Sch. Bip.) with 

 purple pappus is not uncommon from Missouri to Louisiana 

 and Alabama. 



2. Elephantopus tomentosus Linn. 



Sp. PI. 814. 1753. 



As the type of this species, is taken that form having short 

 ovate or elliptical radical leaves. Canby's No. 62 (1898) 

 from Alabama and Pollard's No. 1048 (1896) from Missis- 

 sippi may be taken as very typical examples, though it is 

 common from New Jersey to Texas. There is, however, a 

 very great amount of variation in the vestiture and in the 

 form of the floral and radical leaves. In some the radical 

 leaves are very much narrowed proximally and it seems prob- 

 able that the name nudicaulis Poir. applies to such varieties. 

 Kearney's No. 738 (1897) from Tennessee and Heller's No. 

 77 (1890) from Pennsylvania well illustrate this form. The 

 name nudicaulis Ell. apparently applies to the same thing. 

 It was at first supposed these were the same as elatus Bertol., 

 but for nudicaulis the involucral bracts are spoken of as 



