6. In areas where teosinte does not grow wild (north 
of Mexico, for instance), it should prove relatively easy 
to establish the presence of maize in the pollen record. 
In these same areas, there is no present evidence of wild 
maize in the archaeological or palynological record. 
Theretore, where maize pollen is present in these areas, 
we can presume that the plant was cultivated. How early 
this occurs in the United States and where it occurs are 
important archaeological problems as yet unsolved. 
In conclusion, in identifying gramineous pollen as 
maize, size measurements of large grains should first be 
made. ‘Those grains which fall in the maize- 7ripsacum 
range should then be examined by means of phase- 
contrast optics. Positive discrimination can then be made 
between Zea (maize and teosinte) and T'ripsacum. Fol- 
lowing certain observations outlined in this paper, maize 
grains can be separated from teosinte with reasonable 
reliability. In certain geographic areas, such as the 
United States, grains in the Zea size range with Zea 
rn 
pattern may be presumed to be maize. 
ACKNOWLEDGMENTS 
The authors are indebted to the following individuals 
for aid in the project: Professor P. C. Mangelsdort: 
Protessor W. C. Galinat, University of Massachusetts, 
for much help, criticism and for providing access to the 
resources of the Botanical Museum of Harvard Univer- 
sity ; Professor Donald R. Whitehead, Williams College, 
for critical evaluation and for the data presented in Text 
Figure 1. Both Professor Galinat and Professor Mangels- 
dort have read the manuscript and offered suggestions, 
while the latter has been particularly helpful in deserib- 
ing the relative evolutionary position of the various races 
of maize and teosinte studied. 
[ 44 
