118 A TEXT-BOOK OF GRASSES 
of Eragrostis cilianensis (EH. megastachya) may be consid- 
ered (Fig. 51). The lower pair of bracts are empty and 
are called glumes. The lower is the first glume; the upper 
the second glume. The midnerve is the keel. The bracts 
above the glumes are regularly arranged on a slender 
axis or rachilla, alternately in 2 ranks. These are called 
lemmas. In the axil of each lemma, except 1 or more of 
the reduced uppermost, there is a flower, and between the 
flower and the axis a second smaller bract called the palea. 
The lemmas are also keeled and have a pair of lateral 
nerves. The palea is 2-keeled, and is inclosed within the 
margins of the lemma. The lemma and palea together 
with the inclosed sexual organs are called the floret. 
The spikelet may be interpreted as a specialized 
branch, bearing a series of bracts, or modified leaves, the 
lower pair (glumes) being empty, the others (lemmas) 
bearing a much specialized branchlet (the flower) in the 
axil, the palea being the prophyllum. 
The terminology here adopted differs somewhat from that in 
common use in early works on agrostology. The terms with which 
the student is more likely to come in contact are: For glume, the 
terms empty glume and empty scale; for lemma, the terms flowering 
glume, flowering scale and lower palea or palet; for palea the altered 
spelling palet. The objection to the term scale is that this name is | 
applied only to modified leaves on a shoot below the foliage-leaves. 
The modified leaves of the spikelet are, then, to be included under 
the general designation, bracts. It seems desirable, however, that 
special terms be adopted for these parts as they are so often used. 
In this the writer is following the leading contemporary works on 
plant morphology. Again, for the sake of brevity and precision, 
there is a distinct advantage in using separate terms for the bracts 
containing flowers, and for the empty pair at the base of the spikelet. 
Morphologically the glumes and lemmas are equivalent, and hence 
the terms empty glume and flowering glume are entirely proper. 
But the constant difference in relation and function justifies the 
