Chase — Notes on Genera of Panicese. II. 9 



Trinii Kunth because of P. rigidifolium (Poir.) Kunth (May 2,* 1829, 

 Rev. Gram. 1: 37). Both names thus date from 1829, and we have not 

 been able to ascertain which is the earlier. An older name than either 

 may apply to this species. Doell (1877, Mart, Fl. Bras. 2 2 : 236) uses 

 Panicum dicecum Spreng. (1825, Syst. 1 : 322) for this species, citing 

 " Panicum melicoides Nees ab Esenb. secundum specimena authentica, rix 

 Poiret." Sprengel's description is very brief, and does not seem to apply to 

 this species; the spikelets are said to be dioecious and the leaves lanceolate. 

 The native country is indicated as unknown, and "P. melicoides et pofe- 

 forme Poir." are cited as synonyms. Nees (1829, Agros. Bras. 191) 

 describes this species under Panicum melicoides Poir., giving Panicum 

 diozcum Sprang, as a synonym. Since Nees might be expected to have 

 seen Sprengel's specimen it may lie that Sprengel's name belongs to this 

 plant, notwithstanding the inapplicable description. The original descrip- 

 tion of P. melicoides Poir. (1816, Encyc. Suppl. 4: 283) could hardly apply 

 to this species. Neither the type of this nor of P. dioecum has been seen. 



Two other Brasilian species, Hymenachne campestris Nees (of which 

 Panicum camporum Kunth is a typonym) and P. caudatum Salzm. ; and 

 Panicum interruptum AVilld., of India, and a number of other old-world 

 species of this group are not well enough known to us to be transferred 

 here. 



' Fide Sherbom and Woodward in Journ. Bot. 39 : 2U5. 1901. 



