in deep cavities along the rachis. The general appearance 
is superficially similar to that of 77ipsacum floridanum. 
There are so many differences between T'ripsacum and 
Monerma, however, that there can be no doubt that the 
resemblances are a result of parallel evolution. The rachis 
cavity of Monerma is closed by the second glume (instead 
of the first), is longer than a rachis-segment (instead of 
confined within it), lacks hairy notches on either side of 
the outer glume and lacks the nodal parenchyma which 
connects the rachis-segments in T’ripsacum. The outer 
glume of Monerma, as well as the associated rachis 
groove, extends up into the next rachis-segment and, in 
this way, holds the spike together during the final stages 
of maturation. There is a scar which projects upward 
from the apex of the cavity to the node above. This scab 
or scar is identical to the surface of the cavity below in 
being indurate and in having a glossy-yellow color, as 
compared to the surrounding rachis which is herbaceous 
and green. This depression does not appear to be asso- 
ciated with a prophyll-like structure and is probably me- 
chanical in origin. At least compression of the spikelet 
against the rachis-segment has caused suppression of the 
inner glume and apparently also suppression of any ad- 
jacent prophyll primordia. 
Other grasses. Other examples of rachis cavities, ob- 
viously of a mechanical origin, are found in isolated 
species (Paspalum fluitans (Ell.) Kunth and Stenotaph- 
rum secundatum (Walt.) Kuntze). 
Discussion 
The cupulate rachis-segment, which is basic to the type 
of fruit case in teosinte and T'ripsacum, appears to be, 
in part, the ultimate product of several evolutionary 
trends affecting the inflorescence and, thereby, the pro- 
tection of the caryopses. These general tendencies have 
[ 285 ] 
