Chas< — Notes on Genera of Panicecz. IV. 



117 



(longer than in the other spikelets) and lias much the appearance of a 

 hirsute prolongation of the rachis. The characters to which we here give 

 chief weight as generic are the strictly racemose inflorescence, the paired 

 spikelets t'accti i face (exactly the opposite arrangement to that in Thrasya), 

 the lower or secondary spikelet of the pair ahortive or usually so, the 

 fertile spikelet with its hack to the axis (as in Brachiaria), the long first 

 glume, and the fruit less indurated than in Panicvm. 



Raceme solitary, terminal on the culm or branches; second glume 

 boat-shaped only . . . . . v E. inflexa. 



Racemes several, erect, distant along a common axis; second glume 

 becoming ventricose and uncinate-spiny at maturity . . E.polystachya. 



Echinolaena inflexa (Poir. ). 



Cenchrus inflexus Poir. Encyc. Suppl. 6 : 50. 1804. "M. Richard l'a 

 recueillie a Cayenne. ( V. s. in. herb. Jussieu & Lam. )" The type speci- 

 men in the herbarium of the Museum at Paris was examined by Prof. 



Hitchcock and found to be the same 

 species as the type of Echinolaena 

 hirta Desv. in the same herbarium. 

 Trinius (Gen. Pan. 75. 182(5) also re- 

 ferred C. inflexus Poir. to E. scabra 

 H. B. K. Poiret observes that he finds 

 no involucre, that the species may be- 

 long to another genus, and that it 

 deserves further study. 



Cenchrus marginalia Rudge, PI. 

 Guian. 19. pi. 25. 1805. No locality 

 nor specimen is cited. The type has 

 not been examined but the plate and 

 de : cription identify the species. 



Echinolaena hirta Desv. Journ. de 

 Bot. Paris 1 : 75. 1813. (See above.) 



Echinolaena scabra H. B. K. Nov. 

 Gen. & Sp. 1 : 118. pi. 38. 1816. "Cres- 

 cit in ripa umbrosa Atabapense juxta 

 vicum San Balthasar et rupem Kema- 

 rumo. ( Prov. de la Nueva Guayana. )" The type specimen has not been 

 examined but the plate and description identify the species. 



Panicum echinolaena Nees, Agrost, Bras. 128. 1829. Nees divides the 

 entire species into three varieties, a, p and 7 and cites " Echinolaena 

 scabra ,H. et K." as a synonym without indicating to which variety he 

 considers it synonymous. Since he also cites specimens we can not con- 

 sider P. echinolaena a change of name only. " Var. a et /3 in Brasilia 

 meridional!. (Sellow.) Vidi in Herb. Reg. Berol. )— Var. 7 in campis 

 siccis ad Tanbate nee non Sahara, provinciarum S. Pauli et Minarum 

 * * * (Martins)." The specimens referred to as a and /3 have not 

 been examined. The Martius specimen, the type of var. 7, in the Munich 



Fig. 



Echinolaena inflexa. 

 ( Spikelet and fruit x 6% diam.) 



