PANICE^. 83; 



instead of being marginal on one side of the rachis, they are deeply 

 embedded in alternate cavities on the outer and lower side of that 

 rachis. 



Anastrophus, the third section, was proposed as a genus by 

 Schlechtendahl. In this the spikelets are on alternate margins of 

 the narrow, somewhat flexuose rachis of the spike and the back of 

 the floral glume being turned outwards from the rachis. The 

 spikes are generally several, close together at the end of the 

 peduncle. 



Besides the above valuable notes taken from Bentham, con- 

 siderable aid has been received from a valuable synopsis of the 

 genus by Yasey, in Bull. Torr. Club, 13: 162 (1886). 



A. Anastrophus Schlecht. Spikelets almost distichous with 

 ■ {back of the floral glume) first glume turned outward away 



from the rachis. Leaves obtuse. 



a. Leaves 4 mm. wide or narrower 1 



a. Leaves 5 mm. wide or wider 2 



B. EuPASPALUM Benth. Spihelets strictly secund with {back 

 of the flo7-al glumes) first glume turned inward toward the 

 rachis. Leaves sharp-pointed. 



a. Rachis thin, folding towards the small spikelets and 



partially enclosing them (o) 



0. Spikes 3-6 3 



o. Spikes 3-8 4 



o. Spikes 40-60 5 



o. Spikes 75-100 6 



a. Eachis not enclosing the spikelets, usually less than 1 



mm. wide (b) 



b. Rachis not winged 7 



b. Rachis more or less winged (c) 



c. Spikelets more or less hairy or puberulent. . .(d) 



d. Spikelets 3-2.5 mm. long 8 



d. Spikelets 1.5 mm. long 9 



c. Spikelets glabrous (e) 



e. Spike single at the apex of the culm or each 



branch ; spikelets 1 mm. long 10 



