798 NOTES FROM THE BOTANIC GARDENS, STDNEY, NO. XVI., 



TIii.s slender grass looks very distinct from the robust typical' 

 P. semialatum, with long narrow leaves and spikelets crowded on 

 the panicle-branches, but it has the chief characters of that 

 species, i.e., the densely fringed second glume, and the cleft palea 

 of the third glume. The absence af the nerves on the second 

 glume is caused by the different texture of the glume; the nerves- 

 can be seen also in this variety, if one looks through the glume 

 against the light. 



A remarkable form of the typical robust Fanicioiu semialatum, 

 with broad marginal wings on the second glume of the spike- 

 lets, has been collected by Mr. J. E. Hadley, at Warialda, N.S.W., 

 in April, 1908, but the character is too inconstant for a distinct 

 variety. Normal plants and plants with winged glumes grow 

 side by side in the locality, and in other specimens spikelets with 

 winged glumes and spikelets, or with unwinged glumes are mixed 

 in all proportions. This character is not mentioned by Benthani, 

 in the " Flora Australiensis," and we propose to amend his des- 

 cription so as to include the Warialda form. 



Bentham writes(Vol. vii., p.472) : — "Glumes , the 



2nd the largest, membranous, 5-nerved, fringed on each side with 

 long pale or dark coloured hairs connected at the base on the 

 intramarginal nerve; 3rd glume with a small palea." 



To the description of the 2nd glume should be added : — 

 occasionally winged by a horizontally striate wing fully ^ line 

 broad, and densely ciliate; and in the description of the .^rd 

 glume it should be mentioned that the paiea is deeply cleft. 



Panicum semialatum is also an Asiatic grass, and a form, very 

 closely resembling the Warialda form, is figured by Griffith in 

 his " Icones Plaiitarum Asiaticaruin "(Vol. iii., t.l45), under the 

 name Paniaim viaticum. The second glume in the Asiatic form 

 is broadly winged, as in the Warialda form; the sole difference 

 is that the wing is present only in the lower three-quarters or 

 four-fifths part of the glume, leaving the margins of its upper 

 part bare. 



Unfortunately Griffith published no description of his F. viati- 

 cum, and we have no means of knowing whether he deliberately 



