Caves in Arizona (14); La Perra Cave in northeastern 
Mexico (34); Swallow, Tau, Slab, and Olla Caves in 
northwestern Mexico (88); Cebollita Cave in New Mex- 
ico (Galinat unpublished). Additional collections not yet 
studied but obviously containing tripsacoid cobs have 
been received from Mr. Herbert Dick from a site in 
Colorado; from Mr. Dick Shutler from sites in Nevada: 
and from Dr. Robert E. Bell from a site in Oklahoma. 
Highly tripsacoid cobs can also be recognized in a photo- 
graph of prehistoric specimens from ‘Tularosa Cave in 
New Mexico published by Cutler (9) and in the illustra- 
tion of specimens from the Hueco Mountain Caves in 
Texas published by Cosgrove (cf. 14). 
That these tripsacoid cobs are the result of teosinte 
admixture, which probably occurred in northern Mexico, 
and not of parallel mutations is strongly indicated by the 
fact that some of them have more than one character of 
teosinte, such as distichous spikes, single spikelets and 
highly lignified rachises and glumes. Simultaneous muta- 
tions producing all of these tripsacoid characters are diffi- 
cult to imagine, but genetic recombinations involving all 
of them are common in segregates from maize-teosinte 
hybrids. Both Mangelsdorf (28) and Rogers (48) have 
shown that there is genetic linkage between all of these 
characteristics. 
Perhaps the strongest evidence that these prehistoric 
tripsacoid cobs are the product of admixture with teo- 
sinte lies in the fact that virtually all of them can be 
matched quite closely, sometimes almost exactly, with 
modern specimens derived from experimental maize- 
teosinte hybrids. Galinat ef a/ (14) have illustrated a 
number of these matched pairs, and we have many 
others. Until it ean be shown that there are other and 
better ways of synthesizing facsimiles of the tripsacoid 
prehistoric specimens, we shall continue to assume that 
[ 398 ] 
