tion. The presence of rudimentary spathe initials sub- 
tending each pair of spikelets is revealed by certain mu- 
tant genes of maize (the corn-grass and teopod genes) 
which stimulate them into active development. Also 
these genes may cause development of a prophy]l at the 
base of the spikelets when cupules are absent and, there- 
by, complete the phytomer (Galinat, 1954 0d). 
The fruit case of Coiv, an Oriental relative of maize 
studied morphologically by Weatherwax (1926), repre- 
sents the ultimate in functional development of the spathe 
part of a floral phytomer. In this grass the subtending 
spathes are well-developed as protective enclosures for 
individual pistillate spikelets, and a free prophyll, with 
its dorsal side adaxial to a slender non-cupulate rachis- 
segment, is borne at the base of each spikelet and within 
the spathe. 
EVvoLuTION PRECEDING CUPULE FORMATION 
At the other extreme of the grass family, in certain 
Bambuseae, the floral phytomers may also be completely 
developed. In the bamboo genera of Bentham’s sub- 
tribe I (and also in Nastus), Holttum (1956) found that 
‘the branching of a spikelet tuft is exactly on the same 
pattern as the vegetative branching at the node of a 
bamboo culm’’ [in having a subtending bract and a pro- 
phyll at the base of every branch]; and he suggests fur- 
ther that in the more evolved grass panicles ‘‘the protec- 
tive function of bracts and prophylls... is no longer 
needed [because of protection provided by leaf sheaths 
to the young inflorescence], and the disappearance of 
these structures is to be correlated with that fact.”’ 
It is probable that compression from tightly binding 
leaf-sheaths about young paniculate inflorescences was 
originally responsible for certain reductions and modifi- 
cations within the floral phytomer. For example, the 
[ 227 ] 
