I48 BOTANICAL GAZETTE. [June, 



C. rosea Nutt. Achenia linear-oblong, less than a line 



long, thinnish, almost straight, slightly ribbed on inner face. 

 (Fig. 10.) 



Drawing from Britton's Long Island collection of 1872. I 

 have transferred C. rosea so as to follow C. tinctoria. Al- 

 though it has characters which relate it to C. nuda, for our 

 purpose it seems best to consider it in this section. The fruit 

 and habit are so similar that it seems unnatural not to con- 



sider it as belonging here. 



Achenia oval to obovate, 



half a line in diameter, thickish, wingless but with a cartil- 

 aginous margin. (Fig. 11.) 



The achenia are unlike any others of the genus, but curi- 

 ously enough the species has been much confused in collec- 

 tions. Drawing from Reverchon's collection, distributed as 

 C. tinctoria. 



2. Achenia not villous-ciliate, with a large callus generally devel- 

 oped at each end of inner face. 



Wings thin and broad; pappus two small teeth. 



C. coronata Hook. Achenia oval, 1 to 2 lines long, smooth 

 or tuberculate ; wings from half to as broad as body ; pappus 

 two short or prominent teeth. (Figs. 12 and 13.) 



Considerable variation is found in the size of the achenia, 

 breadth of wings, and length of pappus. Fig. 12, from 

 Drummond's collection, has the wings narrow, as described 

 by Torrey & Gray ; while fig. 13, from Hall's collections, is 

 the more common form. 



liar 



itjL 



fubescens Ell., can not be separated upon 



fruit characters with the specimens at hand. Collectors have 

 almost universally neglected to collect them in fruit, and our 

 material has largeh been obtained from old collections. Dr. 

 Crrav considered that the species were probablv confluent, 

 and the nearly identical achenia and similarity of foliage 

 seem to confirm this. Especially is this true of C. grandi- 

 fiora and C. lanceolata. C. pibescens Ell. in the Syn. Flora 

 is said to have (like C. lanceolata) the pappus very small or 

 obsolete (as in fig. 15), but Curtiss 1485 has awns as long as 

 those of Q, granditlora Fig. 14 is C. grandifiora from Bige- 



lows collections in i8«, Fort Smith, Ark. Fiff. is is C. 



M 



