OF ARTS AND SCIENCES. ' 65 



(Asclept'os vuninolis, Swartz) appears to be the most proper name for 

 the Sarcostemma Brownii (not Bi-oivnei) of INIt-yer and Grisebach, 

 S. claitsnm of Decaisne, mainly, and /S*. crassifolium of Chapman 

 P. undulata (^Sarcostemma unduhitum, Torr. 1. c.) is our only species 

 with a conspicuous column, somewhat longer than the tumid scales of 

 tlie stamineal corona. 



ASCLEPIAS, &c. In elaborating this genus and its allies for the 

 North American Flora, the limitation of the genera has to be consid- 

 ered. More than thirty years ago this work was done by the excellent 

 Decaisne for De Candolle's Prodromus, evidently in a hunied way ; 

 Dr. P^ngelmanu and the late Dr. Torrey have |)ublislied important 

 details; and the former has supplied me wiili full notes and many 

 sketches of his elaborate studies, which, unfortunately, have for a long 

 time remained unpublished. For his I'erent elaboration of the whole 

 order in the new Genera Plantarum, Mr. Benthain could not critically 

 examine all the species. Had he done so, he would probably have 

 either re-established Nuttall's Anantherix (as Dr. Engelmann in his 

 notes long ago proposed), or he would have remanded it to Asclepias. 

 The process of the hood, s[)ecially ciiaracteristic of this last genus, 

 although generally corniform, not vei-y rarely takes the form of a 

 pointless crest or plate, like that which in Ananthertx divides at least 

 the upper part of the cavity. Then Acerates angustifoUa (Polyotus, 

 Nutt.), as Nuttall intimated, has the technical character of Asclepias 

 in a reduced form. I think I have found a character in the anthers, 

 which may be turned to useful account; and it may be of no disadvan- 

 tage that, while distinguishing Acerates from Gomphocarpus (the latter 

 sufficiently heterogeneous without the addition of the former), it helps 

 to separate Nuttall's original Auantherix (the anomalous Asclepias 

 connivens of Baldwin) from the two more common sj)ecies which 

 Nuttall long afierwai'ds added to it, along even with Podustigma, 

 his Stylandra. 



I venture to rearrange the genera in question, in the manner pro- 

 posed in the foot-note.* 



* SYNOPSIS. 

 A. CucuUi coronae basilares intus nmli, ab antheras longe remoti. 

 PoDOSTiGMA, Ell. Coroilae lobi campanulato-erecti. Colunina staminea 

 praelonga. Antherarum alse triangulares basi lata truncata. 



B. CucuUi (aut basilares aut in columna brevi pi. m. elevati) antlieris prox- 

 imae : corolla rotatopatens vel reflexa. • 



* Cuculli intus processu dorsali vel subbasilari corniformi seu cristseformi aucti. 



VOL. XII. (n. S. IV.) 5 



