turity, an extreme induration of the outer glume and 
rachis segment followed by rachis disarticulation with 
basal callus. The outer glume of T'ripsacum and maize 
is smooth, while it is usually sculptured in native New 
World species of Manisuris, the exceptions being MW. 
aurita of the New World tropics, and apparently certain 
variants of M. cylindrica. In any case, the smooth glume 
feature could have been derived as a dominant feature 
from maize. 
Some vegetative features of Tripsacum, such as its 
broader leaves or capacity to evolve broader leaves than 
Manisuris, must have come from maize, while other 
vegetative features are better ascribed to Manisuris. The 
profuse tillering and perennial habit of growth in Trip- 
sacum would probably have come from Manisuris. There 
is no evidence that the mere doubling of the number of 
chromosomes would convert an annual such as maize into 
a perennial such as Tripsacum. Certainly the perennial 
character would have been initially useful in the 2n hy- 
brid by allowing it to persist despite sterility until a fertile 
4n amphidiploid could arise through somatic doubling. 
Assuming, then, that the perennial habit did originate 
with Manisuris (because the oldest archaeological remains 
of maize from Tehuacin, Mexico, indicate that wild 
maize, like modern maize, was an annual (Mangelsdorf 
et al, 1964) ) it must have been strongly dominant in its 
expression. All known New World species of Manisuris 
are perennial, and the rhizomatous habit is especially 
strong in M. rugosa, in which underground shoots may 
extend horizontally for several feet. 
Even more significant than the general intermediate 
position of Tripsacum is the fact that when we use the 
technique of Anderson (1949) in asking the question of 
what source among grasses now known could make the 
necessary modifications in an assumed parent (Manisuris) 
[ 298 | 
