appearance of the ears alone to sort out varieties which 
later proved to have low knob numbers. 
More detailed studies, the results of which are set forth 
in Table II, show that knob number is associated with 
many other characteristics. Before considering these re- 
lationships it should be explained how they are deter- 
mined. Since the data are not extensive in a statistical 
sense and in some cases represent qualitative rather than 
quantitative scoring, elaborate statistical treatment is not 
indicated and in some instances not feasible. P values, 
which are a measure of significance, were derived from Chi 
square values, computed from fourfold tables in which 
the population was arbitrarily divided into two groups on 
the basis of knob numbers and into two additional groups 
on the basis of the other character under consideration. 
In several cases where additional refinement seemed de- 
sirable the second division involved three categories in- 
stead of two. 
In quantitative characters other than knobs, separa- 
tion was made at the mean and the two groups were ap- 
proximately of equal size. In the case of knob numbers, 
however, it was found by examination of the data and 
preliminary computations that a more natural grouping 
could be made if varieties with six knobs or less were 
included in the low-knob group, varieties with seven 
knobs or more in the high-knob group. P values are only 
slightly higher however, and in several instances lower, 
if the separation is made at the mean. 
RELATIONSHIPS BETWEEN KNOB NUMBERS AND 
OTHER CHARACTERISTICS 
Altitude 
Adaptation to altitude may well be one of the charac- 
teristics involved in the introgression of 'Tripsacum into 
maize. At any rate there is an obvious relationship be- 
[ 224 | 
