ured on the third secondary branch of the tassel, from 
the central axis to the base of the first spikelet. Tripsa- 
coid influence greatly condenses this distance, although 
adverse conditions may also do this. 
Glume shape is correlated with grain shape as Ander- 
son suggested (1944). In Tripsacum and ‘Tripsacoid 
maize the glumes are broad at the tip, not tapered to a 
point as in those plants considered as pure maize, Just as 
grains of Tripsacoid maize are widest near the tip, often 
flattened across the top. 
Anthers vary in shape and size, but no good correlation 
with any tendency or within any race could be discovered. 
This is also true of pollen. The maize which is considered 
most contaminated by other grass species has small pol- 
len, just as one would expect because practically all other 
grass pollen is much smaller than that of maize. The pol- 
len of other maize (sweet corn, for example) is occa- 
sionally small. 
The arrangement of paired spikelets on the tassel 
branches is highly indicative of tendencies which are 
present in the plant. One of the spikelets is usually ped- 
icellate while the other is sessile, a condition which is not 
only common in grasses but is found in the Palmaceae 
and many other plant families. The arrangement of these 
pairs on the branches is nearly always regular (Figure 
2 A) and any variant follows certain forms. An index of 
the divergence from regularity of arrangement has been 
devised by Dr. Edgar Anderson (1944) and called con- 
densation index. His definition, ‘‘The average number 
of spikelet pairs per apparent node in the most condensed 
central three quarters of the basal-most secondary 
branch,”’ is applicable to most North American tassels, 
but when one studies the arrangement of spikelets of all 
maize, it is apparent that condensation index is increased 
in two distinct ways, by condensation and by multiplica- 
[ 268 ] 
