sheath, whereas the portion which is adjacent to the blade 
is reduced and papery. A similar phenomenon occurs in 
the lemmas of many grasses (4vena and Bromus) where 
the blade is modified into a narrow awn. 
The effect of the /¥@ gene in reducing the size of the 
cupule wings is not necessarily comparable to its effects 
on either glume wings or ligules. Rather the small size 
of Ie spikelets may decrease their capacity to depress 
the adjacent rachis-internode during youth and thereby 
result in shallow cupules. 
Discussion 
The features of Tu and V2 glumes differ from normal 
in Opposite ways and in doing so they reflect glume-ty pes 
involved in the early evolution of maize and T'ripsacum 
in the American Jaydeae. Teosinte is thought to have 
developed later from the hybridization of maize and 
Tripsacum (Mangelsdort and Reeves, 1939). The folia- 
ceous character of tunicate glumes is typical of the An- 
dropogonaceous grasses (as in HNlyonurus tripsacoides). 
Also it is a starting point from which the glumes of 
maize, Tripsacum, and possibly Manisuris could have 
evolved (Galinat, 1956). In the differentiation of T?ip- 
sacum and Manisuris, the glumes underwent reduction 
and lignification or sculpturing while the paired grain- 
bearing spikelets were reduced to singleness. Maize 
evolved in a separate direction which was controlled 
largely by mutation at the Z'w-tw locus (Mangelsdorf, 
1948). Nevertheless, a latent ability to evolve a Tripsa- 
coid-type of outer glume might have been retained in 
modern maize and finally expressed as the /g mutant. 
We conclude that in regard to this series (7; normal; 
Vg), the structure of tunicate glumes indicates that the 
Tu-tu locus may have been involved in previous evolu- 
tion of the maize glume and that the mutation to the 
