A Revision of the Western Australian Species of Triodia R.Br. 21 
the sheaths. At the orifice of the sheath there are long hairs on the auricular 
ridge and these are usually matted together with the resin. The ligule is 
a row of short hairs extending right across the inner face of the top of the 
sheath. The lamina is narrower than the sheath. The petiole-like base is 
shorter than in most species. The margin of the lamina is scabrid and the 
apex pungent but the point is not rigid as in fcl Buck Spinifex.” Panicle 
pyramidal at anthesis but in most herbarium material the branches erect 
and the spikelets clustered together, variable in size and from 10-40 cm. 
long. Spikelets linear to ovate (in spikelets with fewer lemmas), with more 
or less imbricate lemmas, shortly pedunculate or almost sessile along the 
panicle branches, 4-11 florets. Glumes lanceolate, ovate or oblong, concave, 
becoming indurate ; nerves 5-7 usually obscure ; glabrous or minutely 
scabrid ; the apex acute, shortly aristulate or ragged. Lemmas divided into 
an entire, indurate basal portion which covers the floral organs and three 
rigid, erect or spreading acuminate lobes which vary from as long as to longer 
than the basal part ; the latter apparently without nerves, pubescent at the 
base and up the middle of the abaxial surface ; the lobes with 3-5 nerves 
each bounded by a green strip of chlorenchymatous tissue, and a thin scarious 
margin which is minutely ciliolate. Palea elliptical, slightly longer than the 
entire portion of the lemma and usually curved over the floral organs, apex 
ciliate, nerves narrowly winged. Anthers oblanceolate, dehiscing from sub- 
apical slits. Caryopsis oblong. (PI. I., fig 3.) 
Despite the wide variation in the growth form and in the dimensions of 
the parts of the spikelets, the author, after due consideration of both herb- 
arium material and field information has come to the conclusion that varieties 
in the taxonomic sense cannot be distinguished. It has been found that 
spikelet variations cannot be correlated with differences in habit. Thus 
growth forms with quite different values as feed cannot be recognised simply 
from a herbarium specimen. The differences between the majority of the 
material are differences of degree only, e.g., relative length of spikelets, relative 
length of glumes to lemmas, etc. Again, Hubbard’s view (leones PI. iv. ii. 
t. 3336 : 1937) that the western material represents a distinct species, has 
been disagreed with for the same reason. The western specimens have a 
more heavily indurate base to the lemma, which is usually yellow and horny 
but intermediates occur. The spikelets in our material have more florets 
but this has been found to depend partly on the vigour of the plant, which 
is related to the habitat or to the time of year in which the panicle develops, 
which again is a matter of habitat. Apart from field notes more than a hun- 
dred separate collectings were available on which to base the conclusions 
expressed herein. 
Nevertheless the growth forms that are evident in the field are described 
below so that some idea of the variation of the species is made available. 
All these forms and less distinct ones, not described, have their significance 
for the pastoralist. The grazing animals (sheep) show definite preferences 
for some forms. This is a result not only of different food values but also 
because of more direct reasons for palatability, e.g., the leaves are less resinous, 
or less pungent, or less scabrid and fibrous. The habitat effects the form to 
a certain extent though broadly speaking there are few major soil alterations 
throughout the area over which this species is distributed (except 80 Mile 
Beach country). 
The forms are divided into two groups : — (a) forms with a dense cushion- 
Jike tussock which does not develop long runners, {b) forms with a tussock 
formed chiefly by loosely tangled stolons or runners which develop semi- 
independent tussocks at their ends. The former group is the larger. 
