156 THE VICTORIAN NATURALIST. 



December, 1899. Billabong of Wimmera, Lowan ; F. M. 

 Reader. 



Stipa pubescens, R. Br., var. tenuior, new variety. 



A shorter and weaker plant. Empty glumes much smaller. 

 Awn pubescent with short hairs. 



November, 1898. Desert, Lowan ; F. M. Reader. 



Stipa pubescens, R. Br., var. auricoma, new variety. 



From less than a foot to about three feet high. Basal leaves 

 frequently short. Leaves with the sheaths more or less hairy 

 with spreading hairs, or scarcely scabrous. Glumes much smaller. 

 Flowering glume densely covered with shining yellowish- or 

 reddish-brown hairs. Awn shortly pubescent or nearly glabrous. 



November, 1897. Desert, Lowan ; F. M. Reader. 



Stipa aristiglumis, F. v. M., var. cana, new variety. 



Lower leaves with the sheath softly pubescent with rather long 

 and spreading hairs, the lamina rather rough to the touch. 

 Teeth of empty glumes approaching S. pubescens. Awn slightly 

 pubescent. 



October, 1898. Desert, Lowan; F. M. Reader. 



Stipa scabra, Lindl., var. pallida, new variety. 



Culms very short. Sheaths generally loose, the upper one 

 embracing the base of the panicle. Panicle much longer than 

 the culm, contracted and frequently dense. Glumes usually 

 shorter. Awn shorter. 



November, 1898. Desert, Lowan ; F. M. Reader. 



Stipa scabra, Lindl., var. subtricha, new variety. 



Upper sheath embracing the base of the panicle. Awn shorter, 

 pubescent or puberulous below. 



January, 1899. Sandy regions, Lowan; F. M. Reader. 



In Stipa eremophila the lobes of the flowering glume are quite 

 as long as in S. flavescens ; in the latter, however, the ligules are 

 without cilice, and in some varieties of S. pubescens and S. scabra 

 the lobes are also of quite the same length as those of S. elegant- 

 issima and S. flavescens, but shorter than in S. eremophila. 

 Hence the absence or presence of the lobes cannot always be 

 relied upon in discerning species and varieties of this genus, and 

 thus S. eremophila should rather occupy a place near S. hemipogon 

 and S. pubescens than be placed in a new sub-section with the 

 flowering glume lobed and the ligule ciliated. The ciliated or 

 non-ciliated ligules in the respective species appear to be 

 constant, so that it is advantageous, and in order to obtain a 

 ready line of demarcation, to modify the three sections : — 



1. Flowering glume glabrous or slightly hairy (in some forms of 

 S. pubescens (3rd section) the flowering glume is fre- 



