Not altogether consistent with this suggestion is the fact 
that Breve de Padilla, the suspected hybrid, appears be- 
fore Harinoso de Ocho, the putative parent. However, 
the former occurs first in the upper level of the Mesa de 
Guaje phase, the latter in the lower level of Palmillas, 
the succeeding phase. Also, the Harinoso de Ocho had 
so low a frequency at all levels that its absence in the 
Mesa de Guaje phase may represent nothing more than 
a sampling error. 
Whatever its origin, Breve de Padilla appears to have 
been a productive race which rapidly replaced other 
races. Starting with an initial frequency of 4.6 percent, 
the cobs of this race increased in succeeding phases to 
23.5, 85.1, and 43.1 percent. In the San Antonio phase, 
A.D. 1450-1800, this was the predominating race. Alto- 
gether 836 cobs were assigned to this race. 
Cobs identified as those of Breve de Padilla had pre- 
viously been found among the later prehistoric specimens 
uncovered in the Sierra de Tamaulipas caves (Mangels- 
dorf et al., 1956; MacNeish, 1958) but their frequency 
was low. This may be regarded as a further bit of evi- 
dence that this race is western in origin and that La 
Perra and Canyon Infiernillo caves represent different 
peoples with respect to their agriculture. 
Minor Races 
A total of 112 cobs were found which could not be 
assigned to any of the races described above but all 
seemed to be related to others of the existing races de- 
scribed by Wellhausen et al. (1952). The following races 
were identified among the cobs. 
Harinoso de Ocho. (Plate XI, B). As already men- 
tioned Harinoso de Ocho first appeared in the lower level 
of the Palmillas phase, A.D. 200-850. At no time, how- 
ever, did it attain a high frequency and the total num- 
[ 41 ] 
