Hitclicock — The Grasses of Hawaii 215 



ing awns, the hairs at base about half as long; first glume oblong, flat on the back, acute, 

 scaberulous, pilose on the back with several long hairs, the apex bidentate, the teeth short- 

 awned ; second glume keeled, scaberulous toward the tip, about as long as the first, tapering into 

 an awn 5 mm. long; sterile lemma hyaline, about as long as the first glume; fertile lemma 

 hyaline, bidentate, between the teeth extending into an awn 2 cm. long, tightly twisted to the 

 first bend (about 5 mm. ) ; pediceled spikelet about 5 mm. long, the pedicel about 3 mm. long, 

 thick and triangular like the rachis joint, long-pilose on one angle, ciliate on the other angles; 

 glumes similar to those of the sessile spikelet, the first more distinctly nerved, the second shorter- 

 awned; awn of fertile lemma shorter than that of sessile spikelet (fig. 109). 



Open rocky places. Originally described from "0-Wyhee (inss. Sandwich — - 

 in sinu Byronis 11)."' Hackel,'" who received from Maximowicz an authentic 

 specimen, states that it does not well agree with Trinius's description. He there- 

 fore describes the species anew. In the Gray Herbarium is a specimen of this 

 species from "Herb. Soc. Hort. Lond." collected on "Ins. Owhyhee, ad sinum 

 Byron" by Macrae. Macrae visited the Hawaiian Islands about 1825, having been 

 sent out on an expedition by the Hortictiltural Society of London. Trinius's 

 description applies to our plant fairly well, except that he gives the 

 number of spikes as 3 or 4 and states that the nodes are glabrous and the first 

 glume II to 13-nerved. There are also slight discrepancies in some of the meas- 

 ttrements, but altogether the differences do not seem to ht important enough to 

 justify discarding Trinius's name in view of the evidence presented by the authentic 

 specimens. It is highly probable that the specimen cited by Trinius was a duplicate 

 from the Macrae collection, another specimen from the same collection being the 

 type of Hackel's Ischaemiim lutescens. ■ I have therefore taken up Trinius's name. 

 Oahu: Without locality, Remy no (Gray Herbarium). 

 Molokai : Wailau Valley, Forbes 527. 

 Hawaii: Hilo, Hitchcock 14144; Newell in 1917; Wilkes Expl. E-xped. Halawa, 



Faurie 1349. Rainbow Falls near Hilo, Hitchcock 14 194. 

 Without locality: Wilkes Expl. Exped. 



Spodiopogon aureus Hook. & Arn. is admitted to the Hawaiian flora by Hillebrand on 

 the authority of Munro but this reference is probably incorrect. 



46. ANDROPOGON L. 



Spikelets in pairs at each node of an articulate rachis, one sessile and perfect, the other 

 pedicellate and either staminate, neuter, or reduced to the pedicel, the rachis and the pedicels of 

 the sterile spikelets often villous, sometimes conspicuously so ; glumes of the fertile spikelet 

 coriaceous, narrow, awnless, the first rounded, flat, or concave on the back, several-nerved, the 

 median nerve weak or wanting; sterile lemma shorter than the glumes, empty, hyaline; fertile 

 lemma hyaline, narrow, entire, or bifid, usually bearing a bent and twisted awn from the apex or 

 from between the lobes; palea hyaline, small, or wanting; pedicellate spikelet awnless, in some 

 species staminate and about as large as the sessile spikelet, in some consisting of one or more 

 reduced glumes, or glumes wanting, only the pedicel present. Rather coarse perennials with 

 solid culms, the spikelets arranged in racemes, these numerous, aggregate on an exserted 

 peduncle, or single, in pairs, or sometimes in threes or fours, the common peduncle usually 

 inclosed by a spathelike sheath, these sheaths often numerous, forming a compound inflorescence. 



-"■ In De Caiidolle, .Alphonso, et Casimir, Monographic phancroganiaruni. Vol. 6, p. 222. 1889. 



[117] 



