slender rachis-segments of tunicate teosinte differ signi- 
ficantly from those of Elyonurus by the presence of a 
shallow non-functional adnate prophyll. 
The similarities between tunicate teosinte and Hlyonu- 
rus tripsacoides not only reflect the close relationship be- 
tween the Andropogoneae and the American Maydeae, 
but also are suggestive of asingle wild prototype of both 
maize and 7T'ripsacum with its cupulate fruit case. Al- 
though the exact nature of wild maize is uncertain, the 
many recent studies on primitive archaeological maize 
suggest that the factors for an extremely compact, poly- 
stichous and continuous rachis represent acquisitions 
made chiefly during domestication. Other peculiarities 
of maize, such as the development of both pistillate spike- 
lets of a pair rather than the complete abortion of one 
and the lack of sculpturing or extreme lignification in 
the outer glume, seem to place wild maize at an evolu- 
tionary point preceding both T'ripsacum and Manisuris 
and similar to a type of Hlyonurus with perhaps the ad- 
dition of an adnate prophyll. Because of the uncertainty, 
however, in Plate LVI we have arbitrarily represented 
maize at the extreme end of the American Maydeae in 
order to present a continuous series of types leading to 
the formation of the cupulate fruit case. In order to facili- 
tate the comparisons, the actual representation of maize 
shows a disarticulating type of rachis. Normally maize 
has a continuous rachis except in certain derivatives from 
maize-teosinte hybrids. 
EVOLUTION OF THE OUTER GLUME 
Concurrent with the evolution leading to cupule de- 
velopment, the outer glume has undergone a correspond- 
ing degree of specialization toward affording increased 
protection. The structure of this glume has evolved from 
that of a long, herbaceous, leaf-like bract (Hrianthus spp.) 
[ 2381 | 
