have practically no rind and the vascular bundles are 
widely scattered. At the other extreme are varieties with 
stalk so tough and strong that it is almost impossible to 
cut them with anything short of an ax. In these varieties 
the rind is prominent and the vascular bundles numerous, 
characters which are definitely Tripsacoid. 
INTERRELATIONSHIPS OF CHARACTERS 
Since practically all of the characteristics considered 
here have shown a relationship with knob number, it is 
to be expected that they should also show interrelation- 
ships among themselves. ‘These relationships exist and 
can undoubtedly be demonstrated statistically in many 
eases. It is not within the scope of this paper to examine 
this problem in detail but it should be pointed out how 
frequently certain characteristics are associated in the 
low-knob group. All but two of the varieties with six 
knobs or less are grown only at high altitudes, 6500 feet 
or more. All of them are susceptible to smut; all except 
one have pubescent leaf-sheaths; all except two are sus- 
ceptible to lodging; all but five have non-fibrous root 
systems. In the high-knob group, however, this asso- 
ciation breaks down, although it does not disappear com- 
pletely. Among varieties with seven knobs or more, are 
thirteen which lack pubescence but show fifty per cent or 
more of lodging; while four pubescent varieties show no 
lodging and five additional ones show only ten to twelve 
per cent. A similar situation obtains with respect to the 
pubescence-smut and lodging-smut relationship. 
All of this indicates that the low-knob group is a fairly 
homogeneous one only slightly affected by such Tripsa- 
cum admixture as it has received. This in turn suggests 
that there is a threshold below which the Tripsacum ad- 
mixture has very little effect, at least upon some char- 
acters. After this threshold is crossed, the effect is not 
[ 285 ] 
