HITCHCOCK AND CHASE — NORTH AMERICAN GRASSES. 



67 



Fig. 16. — Cenchrus microcephalus. 

 From the type specimen. 



DESCRIPTION. 



Plants probably perennial, tufted, witb numerous leafy sterile shoots at the 

 base, glabrous as a whole ; culms 30 to 70 cm. tall, compressed, slender, scabrous 

 below the spike, ascending from a decumbent base, branching from the middle 

 and upper nodes ; sheaths, especially those of the sterile shoots, strongly keeled, 

 pilose on the margin toward the summit and on the shoots, with a tuft of white 

 hairs on each side at the apex, this inconspicuous on the old sheaths; ligule 

 ciliate, about 0.5 mm. long; blades folded 

 at base, often flat above, rather thin, mostly 

 10 to 20 cm. long, 2 to 3 mm. wide, pilose 

 on the upper surface ; spikes mostly short- 

 exserted, 3 to 5 cm. long, the slender axis 

 flexuous, scabrous; burs (including the 

 bristles) about 6 mm. long and 5 mm. wide, 

 the body scarcely wider than the thick 

 base, minutely pubescent; spines flat, 

 broadened at base, the lowermost short and 

 spreading, the upper stout, ciliate at the 

 base, shorter than the 5 or 6 lobes of 

 the involucre, these erect or ascending, ciliate nearly to the summit, 

 rigid but relatively blunt; spikelets usually 2, 4 to 4.5 mm. long, about 

 1 mm. wide; first glume nearly half the length of the equal sterile lemma 

 and fruit. 



Known only from the Berry Islands, a second specimen having been collected 

 on Frozen Cay (Britton & Millspaugh 2211). 



11. Cenchrus pauciflorus Benth. 



Cenchrus pauciflorus Benth. Bot. Voy. Sulph. 56. 1840. " Bay of Magdalena," 

 Lower California. The type specimen, collected by Barclay, is in the Kew Her- 

 barium. Doctor Stapf has kindly sent three burs from this collection. He 

 writes that there are two sheets absolutely identical, both bearing, in Ben- 

 tham's handwriting, the name and the locality as published. Two specimens 

 from Lower California, Xantus's no. 115, from Cape San Lucas, and Brande- 

 gee's no. 3 in 1889, from Boca de las Animas, and one from Yaqui River, 

 Sonora, Palmer's no. 11 in 1869, were sent to Doctor Stapf for comparison 

 with plants collected by Barclay. Doctor Stapf writes : " There is no doubt 

 that they are identical." These plants are slender, somewhat depauperate 

 specimens with burs smaller than the average for the species. Unfortunately 

 the type on which the name of this species is based is not typical of the species. 

 Besides the illustration of the bur from the Barclay specimen a bur typical 

 of the species is shown (figure 18). 



Cenchrus roseus Fourn. Mex. PI. 2:50. 1886. "Vera Cruz (Gouin n. 42 part et 

 43) ." The Gouin specimens in the herbarium of the Paris Museum were examined 

 for us through the kindness of the director. The plants are fragmentary, with 

 very few burs. The notes furnished on the specimens place them with little 

 doubt in C. pauciflorus. 



Cenchrus echinatus forma longispina Hack, in Kneucker, Allg. Bot. Zeitschr. 

 9: 169. 1903. "Oxford in Connecticut . . . leg. E. B. Harger," no. 426 of 

 Kneucker's " Gramineae exsiccatae." A specimen of this collection is in the 

 National Herbarium. 



This is the species to which the name Cenchrus triouloides was commonly 

 applied until 1908, when Professor Hitchcock published 1 the results of his 

 study of the grasses in the Linnaean Herbarium, showing the Linnaean species 



'Contr. U. S. Nat. Herb. 12: 127. 1908. 



