18 BULLETIN 1386; U. S. DEPARTMENT OF AGRICULTURE 
retain approximately their original forms except that their margins 
flatten out against the caryopsis. The characteristic mentioned by 
Schumann {14), that of involute glumes, although not common in 
any of the varieties studied by the writer, sometimes occurs in one 
variety/'' In certain of these varieties also the margins are strongly 
infolded at the apices. 
As in the case of the spikelets, glumes may be acute, obtuse, or 
acuminate, and in some varieties also the first glume is often toothed 
at the extremities of the keellike nerves. 
THICKNESS AND TEXTURE OF THE GLUMES 
In all varieties of sorgo at least one glume is always somewhat 
hardened, and in most of them both are considerably thickened, the 
thickening usually being greatest at the base and decreasing toward 
the summit and the margins. The margins, too. are usually some- 
what hardened, and in some varieties they are somewhat thickened, 
but their outer borders are always thin. Frequently they are hyaline 
and so delicate that they almost or entirely disappear by the time 
the seeds are mature. The degree of thickening at the base is more 
or less proportional to the extent thickened toward the summit. 
Differences in hardness and in thickness are easily detectable when 
spikelets are dissected, there being four or five stages or degrees of 
thickness somewhat definitely recognizable in the varieties that have 
been investigated. 
The distance from the apex downward that the glumes will permit 
folding without breaking is used as a partial test for the extent of 
hardening. In some types, however, the glumes are entirely brittle, 
lacking pliability even at the apices, and in such cases their appear- 
ance at the apices, in the margins, and when crushed may be used as 
a partial indication of the relative amount of hardening that obtains. 
The amount of both hardening and thickening is also indicated in a 
general way by the length of the nerves from the apex downward, 
this feature being considered further in a subsequent paragraph on 
venation. 
Tie' amount of thickening is never easily indicated, as there is no 
way of distinguishing between different degrees. In the variety 
represented in Plate XI. G, the second glume is hardly thickened at 
all. and the first is thickened below only. The variety shown at A 
represents the other extreme, the glumes being somewhat thickened 
even at the apices and in the margins. Although such differences 
are manifest between certain varieties, being plainly discernible 
with a slight magnification, they can not be readily measured or 
adequately shown by means of photographs; and although such 
variations in thickening very evidently indicate relationships be- 
tween certain forms, the character is manifestly not a good one for 
purposes of identification. 
3 Occasionally lateral pressure exerted by the ulumes on the caryopsis, resulting from 
the glumes becoming strongly incurved on the sides, has the effect, mentioned by Schu- 
mann, of causing the caryopsis to rotate in the spikelet. This is seen more frequently 
in herbarium specimens than in panicles in the field, and it is doubtful whether it ever 
occurs in the varieties studied except when the grains are dislodged by handling the 
panicles or possibly in the field from agitation by wind. 
