8 GRAMINES. 
tral one is rather faint; it is besides flat, or rather bicarinate, 
the space between the two lateral veins being convex out- 
wards. The remains of the ovarium are very long. Kunth des- 
cribes the palea of the female flower thus, the inferior 3 
nerved, the superior 4 nerved, the two other veius may exist 
in the inferior, but they are very indistinct; but I find it is 
not so in my specimens, and the latter nervature is incompa- 
tible with a central vein, which certainly exists in this, and 
the nearly allied grasses, 
In the female it is highly developed, and forms ап absolute 
cup. 
The spicule of either sex terminate a dilated foot stalk, 
the margin of which is submembranous, is it owing to the 
great development of this, that Kunth describes Zizania as 
having very minute glumes, connate into one orbicular. It 
is quite evident, I find, that he does so, for he describes the 
female flowers of Hydropyrum as glumate! This is a great 
oversight, for if this cup be a glume; Oryza has 4 glumes, 
and Leersia ciliaris, is also glumate. 
Our indian hexandrous grasses are 
l. Leersia ciliaris, Pharus ciliaris. Retz. for I see no rea- 
son why it should not belong to this genus, at least so far as 
its written character is concerned. 
2. Oryza sativa, 
3. Oryza altera from the Khasiya hills. 
4. Sclerophyllum coarctatum, Oryza coarctata, Roxb. 
5. Potamochloa Retzii, Pharus aristatus Retz. 
Oryza triticoides ? Pl: CXLII. Fig. І. O. coarctata Buch. 
am. 
Gramen in aquis salsolis proveniens. Culmis longissimis 
emersis bi-tripedalibus, subsimplicibus vel parce ramosis (ra- 
mis ex axillis superioribus immersis) (in parte immersa ad ge- 
nicular. radicant.) teretibus glaberrimis, nitidis vaginis digiti 
longitudine  subinflatis rigidis, glabris, (margins  over- 
lapping,) ore quasi perfoliatis, ciliato-fimbriatis. Ligula 
rotundata integra ciliata, lamina lineari-ensiforma, pedali vel 
