﻿342 



BOTANICAL GAZETTE 



obsolete, the upper glume (of the lowest spikelet) lying outside of the 

 internode of the axis, both glumes persistent. 



In the discussion following the description, Hackel states that while 

 he had regarded Aphanelytrum as a subgenus of Brachyelytrum only, 

 further study had convinced him that it was a valid genus. He lays 

 particular emphasis upon the character of the inflorescence: 



The primary branch bears, as it appears constantly, 3 sessile lateral spike- 

 lets, the axis extended as a stipelet beyond the uppermost, thus forming a spike, 

 but of so remarkable a structure that I know of no second example among the 

 grasses. Of the two very small glumes which can always be found distinct 

 only on the lowermost spikelet, those of the two upper mostly being wholly 

 aborted, only the lower is found on the same side of the spike axis as the spike- 

 let, the upper being found on the opposite side of the axis. This position is 

 contrary to the conception of this spike as a monopodium, as the spike of the 

 Hordeae, etc., doubtless is; one must rather assume the axis of the spikelet 

 to be the continuation of the spike axis internode below it,' and the internode 

 next above to represent a branch in the same direction as that below put forth 

 from the axil of the second glume, the whole spike, then, forming a sympodium. 

 Whether this admits of another explanation later study of more material will 



doubt. As to the upper members of the spike these can only be understood 

 from analogy to the lower, since in these the glumes are wanting or reduced to 

 minute vestiges. 



When this structure is recognized as a single spikelet, the necessity 

 for assuming it to be a sympodium is obviated; the position of the 

 second glume is seen to be the normal one; and, except for the elongated 

 rachilla joints, the spikelet is seen to be in no way anomalous. The 

 minute vestiges of glumes below the upper florets referred to by Hackel 

 must be the rather prominent callus of the floret. 



Among plants recently received from Colombia were several speci- 

 mens of Aphanelytrum procumbens collected along a trail at 3100 meters 

 by Fratres Apollinaire and Arthur (no. 717). From these specimens 

 the accompanying figure and the following emendation of the generic 

 description are drawn. 



Aphanelytrum Hack, emended 

 Inflorescence a few-flowered panicle, the remote capillary, flexuous, 

 simple or subsimple branches in fascicles of 2-4 or the upper solitary; 

 spikelets on slender flexuous pedicels, perfect, 2 or 3-flowered, articulate 

 above the minute glumes; rachilla joints capillary, flexuous, from half 



