STREPTACHNE R. BROWN AND PHEIDOCHLOA GENUS NOVUM, ETC. 13 
These specimens show : 
(a) Glumes equal or more or less unequal, in the latter ease 
either the upper or the lower glume the longer; all varia- 
tions to be seen in one and the same panicle; the length 
varies from 9 to 12 mm. in 8149, and 9 to 13-5 mm. in 
13459. 
(5) Lemma, column and awn nearly in a straight line (in 
immature spikelets) to variously bent and geniculate with 
respect to one another; lemma shorter than or as long as 
the glumes; awn straight or gently curved; the lemma 
varies in length from 6-5 to 13 mm. (usually) ; the column 
is 4*5-7 mm. long; the (central) awn varies from 9 to 
15 mm. long. 
(c) The lateral awns vary greatly; when well-developed up 
to 4 mm. long, very often greatly reduced or one or both 
entirely wanting. Figs. 9-12 of the plate illustrate stages 
in the reduction of the lateral awns in different spikelets 
from a single specimen; fig. 12 is that described for 
Streptachne stipoides R.Br., while fig. 9 represents the 
state found in the type of Aristida utilis F.M.Bail. 
12. In one and the same panicle of many specimens of particularly 
Blake 8149 and 13459 occur spikelets which match those of 
the type-specimen of Aristida utilis (figs. 2 and 9) and other 
spikelets agreeing with all the available data concerning the 
type of Streptachne stipoides (figs. 3 and 12). 
From the above data there seems no doubt whatsoever that Aristida 
utilis F.M.Bail. and' Streptachne stipoides R.Br. are conspecific. In 
general the longer the panicle the greater the tendency to produce 
prominently 3-awned lemmas, and the smaller the panicle the greater the 
tendency for the greater number of spikelets to have the lateral awns 
much reduced or absent. As a rule the panicle is about as long as the 
remainder of the culm. It is noteworthy that the panicles of the type 
-of Streptachne are according to Bentham 5-6 in. long* while the usual 
length in A. utilis is at least twice this. 
Banks and Solander were at Endeavour River between June 17tli 
and August 5th, 1770, an unfavourable time of the year for collecting 
grasses in this region. I visited the area at the same season (July 30th 
to August 2nd) in 1943 and found Aristida utilis very abundant, but 
the plants had shed nearly all their florets. A search for plants without 
lateral awns was undertaken, and it was evident that a considerable 
variation occurred. Only a very few specimens in over-mature condition 
or with spikelets arrested in development were procurable, and though 
some spikelets have very tiny lateral awns there are none with them 
entirely absent. The type-specimens of A. utilis from the same locality 
were also collected rather late in the season (May). So it is possible 
that none of these specimens gives a true idea of the species. 
The generic position of the species has now to be considered. 
Hughes, l.c., has given sufficient reason for not uniting Streptachne 
with Stipa; the absence of an articulation between awn and lemma is 
exclusive, besides the frequent presence of lateral awns, which was 
* Mrs. Chase writes me that in the United States National Herbarium there is 
a tracing of these plants sent by Miss Hughes, and the panicles measure 23 and 24 cm. 
long, the two lower branches of each 6-5 and 7 cm. long. 
