PANICLE CHARACTERS OF SORGO 13 
of the axis and of those of the branches of the inflorescence are 
variable, this variability resulting largely from environmental in- 
fluences, so that neither is of much value if employed alone unless 
the varieties happen to have been grown under the same conditions 
of soil and climate. Often, for these reasons, it is only when all the 
characters mentioned are considered together along with certain minor 
features which are discussed in subsequent paragraphs that a con- 
ception of the type of the panicle can be accurately obtained. Con- 
trasts in panicle type are, however, very evident between certain sorts 
and are outstanding characters, varieties often being more easily 
distinguished by such differences in the panicle than by any others, 
such differences also being the ones most commonly employed by 
growers in distinguishing one sort from another. 
LENGTH, FURROWING, AND COLOR OF THE AXIS 
The length of the axis as compared with the length of the panicle 
shows considerable variation. It may extend to the summit of the 
panicle or it may be reduced to shorter lengths. It is sometimes 
so shortened that there is but one whorl of primary rays. When it 
is elongated it is attenuate and when abbreviated it is usually trun- 
cate. The effect of shortening the axis on the form of the panicle 
is usually to make it roughly obovate or obconical. In these forms 
the upper branches are usually as long as or longer than those origi- 
nating at the base. 
Varieties differ in regard to length of axis, in some of them the 
character being partty concealed by rigidity of the branches ; within 
the varieties also there is usually considerable variation among the 
panicles in regard to axis length. In certain varieties the axis is 
rarely less than three-fourths as long as the panicle. In certain 
others it is frequently somewhat shorter. In still others it is often 
much shorter; it may be reduced to a single node and it may vary 
from this until it equals the entire length of the panicle. No variety 
of sorgo has been examined in which all axes were greatly reduced, 
that is to say, to less than about one-fourth the panicle length. 
Among the varieties examined, however, three show it frequently 
so, and in each of these, axes of various greater lengths also some- 
times occur. In one variety of sorgo, axes of several lengths are some- 
times seen on the main culm and branches of the same plant. Two 
panicles of one of the varieties showing this characteristic are 
represented in Plate IV, in one of them the axis being elongated and 
in the other much abbreviated. 
This feature has been employed in keys and descriptions by a 
number of writers describing sorghums, including Forskal (7, pp. 
174-175), Arduino (i), Kornicke (P, pp. 294-315), Schumann (7^), 
and Ball (£), which shows that at various times it has appeared as 
an outstanding character. The fact that in all sorgo varieties 
in which it occurs panicles also are to be seen in which the axes are 
prolonged emphasizes the need of examining varieties at least in part 
in the field, and not depending solely on a small number of dried 
specimens. 
Lines which have been self -fertilized for a number of years have 
shown variation in axis length within progenies resulting from single 
self- fertilized panicles. It therefore seems evident that the mode of 
