secured from the lower levels of the Belles Artes core. 
‘These lower-depth grains received the most careful study, 
although an additional 44 large grass-pollen grains from 
the upper levels of the Belles Artes and Madero cores 
were also examined by the same procedures. 
In the course of making measurements, it early became 
evident that a wide range in the size of the pollen grains 
characterizes each of the three genera, a condition possi- 
bly indicative of their unusual genetic variability. In 
Tripsacum, a wide-ranging and common genus of sub- 
tropical and temperate North America, the average for 
the long axes of the pollen grain varied between 41.7 
in 7. maizar and 57.4 in T. australe, with an extreme 
range in the genus of 83.6u to 64.0u. In teosinte, of the 
three forms examined, the average length varied between 
79.34 and 86.44, with extremes of 74.0 and 102.0p; 
whereas in maize the average range fluctuated between 
87.24 (Mexico ‘‘Nal Tel’’) and 122.8 with extremes of 
72u and 141.7. It is evident that the smaller pollen of 
some of the varieties of cultivated maize measured in this 
study fall well within the range of teosinte and close to 
that of the largest 'Tripsacum grains. It is apparent, 
therefore, that size alone cannot be utilized for the criti- 
cal identification of presumed fossil maize pollen, and 
that of its relatives, unless a sufficient number of intact 
grains are available to be measured and plotted on size- 
frequency curves to show the statistical probability of 
one of three possibilities. Owing to the small number of 
individual intact grains from the Mexican borings, how- 
ever, this procedure was not possible. 
Because of the paucity of structural features and the 
undistinctive sculpture pattern of the pollen exines of the 
three genera under consideration, it became necessary to 
attempt some other means of distinguishing the three 
pollen-types. Consideration of the problem led to one 
[ 281 ] 
