26 
BULLETIN 138G, U. S. DEPARTMENT OF AGRICULTURE 
This deciduous or semideciduous tendency in these spikelets some- 
what depreciates all their features for use in varietal identifica- 
tion, and they are so small that they can not be successfully 
studied unless they are magnified. There are, nevertheless, certain 
differences distinguishing them which are so distinct and obvious 
that they merit presentation. The most pronounced relate to their 
length and prominence, the latter resulting from the combined length 
of the spikelet and its pedicel. 
The range of actual lengths in varieties is apparently not changed 
except in minor degrees by environmental influences, but the differ- 
ences in length within the varieties as well as within single panicles 
is sometimes considerable. In the two varieties between which ac- 
tual lengths show the greatest differences, the average lengths of the 
spikelets are 0.18 or 0.20 inch and 0.09 or 0.10 inch, respectively, 
the approximate ex- 
tents of each of the 
two series being shown 
diagrammatically in 
Figure 2. This differ- 
ence in average length 
of spikelets is sufficient 
to constitute a good 
varietal character were 
it not for their small 
size, their deciduous 
tendency, and the va- 
riation that occurs 
within panicles, which 
make comparisons dif- 
ficult. 
In certain varieties 
the sterile spikelets are 
so prominent that they 
give an appearance to 
the panicle distinctly 
different from that of 
some other varieties in which they are hardly visible. Unfortunately, 
there are also varieties in which the prominence of these spikelets is 
intermediate between these two extreme types, there being, in fact, a 
number of transitional types. In certain varieties, however, a pre- 
dominating number of sterile spikelets, including their pedicels, show 
a length greater than that of the fertile spikelets at correspond- 
ing joints of the rachis. In others the combined length is most 
frequently less than the length of the fertile spikelets at corre- 
sponding joints of the rachis. In a third or intermediate class 
those in which the combined length exceeds the length of the fertile 
spikelet and those in which it does not are approximately equal, the 
difference being so small that a contrast is not discernible. Unless 
there is a pronounced preponderance in the number of one or the 
other the characteristic is of little value. 
Fig. 2. — Ranges of size in sterile spikelets of two varie 
ties of sorgo: A, McLean (Sugar Plant No. 0182) ; B 
Sumac (Sugar Plant No. 0181) 
VENATION 
Unlike those of the fertile spikelet the nerves of the sterile spike- 
let always extend the entire length from its apex to its base, and 
