344 Monography of the North American Cuscutinem. 



ger. The ovary is 2-celled and 4-ovulate : but I have never 

 seen more than two seeds, separated by the incomplete dissepi- 

 ment ; and frequently only a single seed ripens. 



Lepidanche Compositarum. 



Stem low, branching ; flowers closely sessile, conglomerate, 

 5-parted ; tube of the corolla nearly cylindrical, longer than the 

 imbricated calyx, which consists of ten to fifteen scales, twice as 

 long as the oblong obtuse spreading or reflexed lobes of the co- 

 rolla ; stamens equal to the limb, exserted ; scales pinnatifidly 

 laciniate, convergent, covering the ovary ; styles twice as long 

 as the ovary with the stylopodium ; capsule globose, enveloped 

 by the scales of the calyx, crowned by the stylopodium and 

 styles, and covered by the remains of the corolla. 



Var. a. SoLiDAGiNis : flowers smaller ; lobes of the limb reflex- 

 ed ; stylopodium half as large as the ovary. 



(9. Helianthi : flowers larger, lobes of the limb spreading ; 

 scales of the filaments united with one another forming a 5-lobed 

 crown in the tube ; stylopodium larger than the ovary. 



This singular plant appears to be peculiar to the western prai- 

 ries. I have observed it since 1833 in wet prairies around St. 

 Louis,* on Solidago, (also on Vernonia, Ch. Geyer,) and Dr. 

 Clapp has found it on Silphium at New Albany, Indiana; the se- 

 cond variety I have gathered on Helianthns since 1838 in similar 

 localities ; flowering in August and September. These varieties 

 may prove distinct species ; but for the present I am unable to 

 distinguish them by more important characters than those given 

 above. 



The flowers are always 5-parted ; the tube is not exactly cy- 

 lindrical, but a little wider at the mouth than at the base, rather 

 obconic. The styles are longer than in any of our Cuscutas, and 

 almost always unequal ; they are inserted on a distinct stylopo- 

 dium, which is larger than in any Cuscuta. The stigma is cap- 

 itate, as in all American Cuscutas. 



* This is manifestly the Cuscuta Americana (from St. Louis) of Hooker's ac- 

 count of Drummond's collections, in the Companion to the Botanical Magazine, 

 I, p. 173; of which it is remarked, that " Some of the specimens seem to have all 

 the flowers abortive, and turned into scales, which are excessively crowded, and 

 form a dense wreath of a pale slraw-color, around the branch of some shrub." 



