HITCHCOCK AND CHASE GRASSES OF THE WEST INDIES. 379 



few weatherworn spikelets on the Wright specimen were too much broken to 

 permit of diagnosis. Brother Leon at our request kindly visited the locality 

 and sent us a specimen of this apparently very rare species. His letter in 

 regard to this collection is of interest : " I visited the Savana San Juli&n in 

 company with Father Modesto Roca of Guanabacoa. The first two days I did 

 not see any kind of Bouteloua [from the broken Wright specimen we had 

 guessed it to be a species of that genus], not knowing on what border of the 

 savanna were the skirts of woods, and the savanna has a circuit of 40 to 50 

 kilometers. At last * * * we went to a small wood (Cayo Gabino) on the 

 northeastern border of the hacienda [of Senor D. J. M. Lamas] and there I 

 saw a tufted grass which I supposed might be the long lost species of Wright, 

 though at first sight it called to my mind the idea of a small Arthrostylidium, 

 having some resemblance to my no. 4853 [Arthrostylidium capillifolium]. I 

 noted with pleasure that it agreed exactly with your description. Unluckily 

 there remained very few flowers. I entered the wood at different places all 

 around but could not find more of it." It is noteworthy that in his field notes 

 Wright called his no. 3894 "Arthrostylidium." 



81. BOUTELOUA Lag. 1 



Spikelets with 1 perfect floret, crowded along 1 side of a narrow rachis, 

 forming short spikes (rarely but a single spikelet to a spike) ; glumes unequal ; 

 fertile lemma rather broad, usually 3 to 5-toothed, commonly mucronate or 

 awned ; sterile lemmas usually with 3 awns ; spikes racemose, often drooping. 



Second floret a trifid naked awnlike rudiment. 



Lateral awns of rudiment much shorter than the central one or nearly 



obsolete 1. B. vaneedeni. 



Lateral awns of rudiment nearly as long as the central one. 



Plants very small; blades not over 1.5 cm. long; spikes less than 5 mm. 



long 2. B. juncea. 



Plants in rather large mats ; blades commonly 10 cm. long ; spikes 2 cm. 



or more long 3. B. americana. 



Second floret well developed, the lemma evident. 



First floret sterile, the second mostly fertile 4. B. heterostega. 



First floret fertile, the second sterile 5. B. disticha. 



1. Bouteloua vaneedeni Pilger in Urban, Symb. Antill. 6: 2. 1909. 



Very slender, tufted ; culms erect, finally producing short fascicled branches ; 

 blades involute ; inflorescence a very slender raceme 5 to 15 cm. long, of numer- 

 ous small spikes of few spikelets. 



Only known from the island of Anguilla, where the type was collected by 

 Boldingh (no. 3512B). 



2. Bouteloua juncea (Desv.) Hitchc. Contr. U. S. Nat. Herb. 17: 343. 1913. 

 Triathera juncea Desv. ; Beauv. Ess. Agrost. 40. pi. 9. f. 4- 1812. 

 Eutriana juncea Trin. Gram. Unifl. 238. 1824. 



Eutriana ledebouri Trin. Gram. Unifl. 238. 1824. 



Aristida secunda Ledeb. ; Trin. loc. cit. as synonym of Eutriana ledebouri. 

 Atheropogon domingcnsis Spreng. Syst. Veg. 1 : 293. 1825. 

 Itriaena juncea Griffiths, Contr. U. S. Nat. Herb. 14: 354. 1912. 



x This genus has recently been revised by Griffiths (Contr. U. S. Nat. Herb. 14: 

 343-428. 1912), who gives a nearly complete synonymy. Here, therefore, only 

 such synonyms are given as are found in West Indian floras and a few others 

 not included by Dr. Griffiths. 



