14 BULLETIN 1386, U. S. DEPARTMENT OF AGRICULTURE 
expression of the factor which controls the character is of such 
a nature that the latter may be expressed either in the elongated or 
the abbreviated type and that intermediate steps may also be present. 
Although panicles on branches are likely to have relatively longer 
axes than those on initial stalks, this is not always the case, examples 
occurring in the self-fertilized lines in which the reverse is found. 
It is to be noted also that in certain years more of one or the other 
of the extreme types has been in evidence, and although observations 
have not been carried far enough to warrant definite conclusions it 
seems probable that environmental conditions influence the relative 
length in panicles both on initial stalks and on branches. 
Varieties show differences in the number and arrangement of 
branches on the axis, its depth of furrowing, color, and pubescence; 
and in the contracted and compact types and also occasionally in 
the spreading types there are frequently one or more nodes at the 
base of the panicle without branches. The presence of furrows is 
believed by Schumann (IJf) to be due to the pressure of branches 
against the axis in the early stages of development. In certain 
varieties the furrows are shallow and in others they are deeper, al- 
though the differences are never very wide. The axis often has a 
tint resembling the color of the glumes of the fertile spikelet more or 
less modified, but more commonly the color is similar to that of the 
peduncle. 
PUBESCENCE OF THE AXIS 
Three distinct conditions of pubescence occur on the axis: 
(1) Fine pilosity, sometimes much appressed, occurring either 
more or less uniformly over the surface or mainly in the furrows; 
(2) hairiness on the ridges, longer than the latter and more or less 
forming a fringe; and (3) a scabrous-hispid condition on the ridges. 
These three types are shown in Plate IX, A, B, and C. Conditions 
more or less intermediate between these are encountered also. All 
varieties are found to be hairy at the nodes, but some are much 
more hairy than others. A fourth condition, which, however, 
is less distinct as it usually blends into the first, is a somewhat 
tufted or barbate condition at the nodes. Certain kinds are essen- 
tially glabrous except for hairiness at the nodes and fine bristles 
or teeth on the ridges, and others are hairy throughout, the axis 
as a rule being more pubescent in the more compact types of pani- 
cles and less so in the effuse types. In certain varieties where the 
nodes are w T ell separated the spaces between are glabrous or nearly 
so, and where they are close together they are hairy. 
BRANCHES OF THE PANICLES 
The actual differences in the length of branches of all ranks 
arising at the same relative positions in the panicle are considerable 
in extreme types, the relative length of branches in four types 
being shown in Plate X. Their length, however, is much subject 
to modification by influences other than germinal. Differences in 
length are therefore never of more than secondary importance as 
varietal characters, and as would be assumed, in intermediate types 
no lines of separation are possible, length being always so variable 
