88 Annals of the Carnegie Museum, 



Family XYRIDACE^. 

 Key to the Species Enumerated. 

 Leaves reaching a length of 10-20 cm. and a width of 8-12 mm.; spikes about 



1.5 cm. long 186. Xyris ambigua. 



Leaves 8-10 cm. long by 1-1.5 mm. broad; spikes 7-9 mm. long. 



187. Xyris longibracleala. 



186. Xyris ambigua Beyrich. 

 Xyris ambigua Beyrich, in Kunth, Enumeratio Plantarum Omnium, IV, 1843, 



P- 13- 



On white sand in the pine-barrens at Los Indies, May 18, 1910, 

 0. E. Jennings, No. 64Q. General Distribution: North Carolina to 

 Florida and Texas, and the Isle of Pines. 



187. Xyris longibracteata Britton & Wilson. 



Xyris longibracteata Britton & Wilson, in Britton, Studies of West Indian Plants, 

 . VIII, Bulletin of the Torrey Botanical Club, XLIII, 1916, pp. 462, 463. 



, "White sand, vicinity of Los Indios, Isle of Pines {Britton, Britton 

 & Wilson 14213)." Britton, /. c. 



Family ERIOCAULACE.F:. 



Key to the Species Enximerated. 



Small, densely-tufted branching plants, up to 3 cm. high, with leaves less than 

 I cm. long and peduncles usually not over 2 cm. long. 



189. PcBpalanthus alsinoides, variety minimus. 

 Larger, with few or no branches, rosulate tufted leaves usually 3-5 cm. long, and 



peduncles 2-3 cm. or more in length 188. Pcepalanthus seslerioides. 



Leaves rosulate-tufted, about 2-3 cm. long; peduncles 5-10 cm. long. 



igo PcEpalanthus androsaceus. 



188. Paepalanthus seslerioides Grisebach. 



PcBpalanthus seslerioides Grisebach, Catalogus Plantarum Cubensium, 1866, 

 p. 224. {Wright, No. 3234-) 



Growing in the white sand of the pine-barrens near Los Indios, 

 May 17, 1910, 0. E. Jennings, No. 338. Reported heretofore only 

 from Cuba. 



This plant agrees very closely with the description of the Cuban 

 plant, although the leaves are shorter, about 3-5 cm. instead of 8 cm. 

 and the peduncles are shorter, being at the longest 13 cm. instead of 

 30 cm. This may possibly be found to be a variety of P. seslerioides, 

 but a larger series of specimens are needed for study before such a 

 decision can be made. 



