the striking resemblances between the prehistoric and 
modern specimens reflect a corresponding similarity in 
their genotypes. 
ARCHAEOLOGICAL EVIDENCE OF IMPROVEMENT 
These prehistoric tripsacoid specimens not only show 
that corn crossed with teosinte centuries ago but they 
also indicate that corn was improved as a result of the 
admixture. In all of the collections of archaeological 
maize in which the lower levels comprise non-tripsacoid 
corn this early corn is small and uniform in type. In 
higher levels, accompanying the appearance of tripsacoid 
types, there is an almost explosive increase in variability. 
This is especially well illustrated in the collections from 
Bat Cave in New Mexico (87) and from Swallow Cave in 
northwestern Mexico (83). This increased variability, 
which involves types both poorer and better (by modern 
corn-breeding standards) than the original corn, can be 
attributed both to genetic recombination and to heterosis. 
This is especially well shown by the data presented by 
Galinat et al (14), based on 433 specimens in which a high 
correlation, 0.859, was found between length of cob and 
estimated teosinte introgression. The correlation is 
strongly curvilinear, both the shortest and the longest 
cobs being highly tripsacoid. This is explained by as- 
suming that the short, strongly tripsacoid cobs are homo- 
zy gous for genes introduced from teosinte, while the long 
cobs are heterozygous for such genes and are the vigor- 
ous products of maize-teosinte heterosis. 
It is improbable that the large modern ear of corn 
could have evolved except for hybridization of corn with 
teosinte which contributed genes for induration and lig- 
nification of the tissues characteristic of the prehistoric 
tripsacoid specimens. On this point Mangelsdorf (29) 
has expressed the following conclusion: 
[ 399 | 
