301 
XLIV.—_GARNOTIELLA. 
Otto Srapr, 
In 1896 I described Garnotiella, a new genus of Gramineae, in 
Hooker’s Icones Plantarum, tab. 2494. The name was suggested by 
its resemblance to and assumed affinity with Garnotia. It was based 
on a specimen collected by Vidal (no. 3994), in the Philippines. 
Some time ago, Mr. E. D. Merrill was good enough to call my 
attention to some specimens in his Philippine collections (no. 4322), 
which evidently represented the same plant and had been named 
Andropogon leptos, Steud., by Professor E. Hackel. The renewed 
examination of Vidal’s specimen and its comparison with Hackel’s 
description of Andropogon leptos left no doubt about the identity of 
the grasses and their position in Andropogon, if this genus is taken 
in the wide sense of Hackel’s conception. This author placed it at 
the end of his sub-genus Sorghum as “species valde peculiaris.” 
Steudel, on the other hand, enumerated it under the section Chryso- 
pogon, and Nees named it, according to Steudel, Chrysopogon 
tener (MS.). As neither my own nor Hackel’s description is 
quite complete or correct, it may be worth while to supplement 
and amend them with a view to establishing more clearly the 
sytematic position and status of the plant. ; 
The mistake J made in comparing the grass with Garnotia was 
mainly due to my overlooking the rudimentary pedicels which are 
hidden away among the hairs surrounding the base of the spikelet, 
and in misinterpreting the awned innermost glume which in mature 
specimens like Vidal’s readily becomes detached. Counting this 
glume as glume iii, to use Hackel’s terminology, instead of as 
glume iv, the structural resemblance of the spikelets of Garnotiella 
and Garnotia becomes quite evident ; moreover, it is supported by 
the general aspect of the inflorescence. : : 
he rudimentary pedicels, that is the barren primary axis of the 
last divisions of the panicle, vary from 0°3-0°5 mm, in length, and 
from 0°03 to 0:005 mm. in diameter, and although—with the ex- 
ception of the hyaline tips—multicellular in cross-section, are more 
or less transparent. They are ciliate, and overtopped by their own 
cilia and the hairs of the callus, some of which attain up to almost 
1 mm. in length. The lower involucral glume has been described 
as nerveless ; but under a high power, and when the glume is made 
sufficiently transparent, two pairs of extremely delicate nerves can 
e observed. They consist in cross-section of one or in the upper 
part of two extremely fine tracheids, about 3u in diameter, and run 
towards the two apical teeth of the glume, leaving a slightly thinner 
middle band between them. The glume is rounded on the back, 
and possibly, under some circumstances, slightly concave co obi 
the nerve pairs as Steudel describes it. The upper involucra 
glume is distinctly laterally compressed and keeled. Flattened out, 
character as in the lower glume ; but there are only three of them, 
one along the middle line, and one on each side, s htly nearer to 
