to the glume supply. In any case, the actual elaboration 
of the glume cushion into a well-developed leaf under 
certain conditions mentioned previously seems to leave 
little doubt about the homology of this rudiment. 
The suppression of the primary leaves along the rachis 
is associated with a loss of the nodal plates and differ- 
ences in trace connections from the axillary buds. ‘These 
traces from the binate spikelets of the rachis fan out to 
the nearest group of ‘‘common’’ bundles (Plate II, fig. 
1) rather than connecting through a vascular network 
extending to the entire meristele, as with the axillary 
buds along the culm. 
Nature of the bundles in the cupule wings. The evi- 
dence from vascular anatomy does not support the sug- 
gestion of Nickerson (1954) that the bundles in the cu- 
pule wings are those of a prophyll adnate to the rachis. 
Although the bundles located near the lateral edges of 
the out-folded wings of a typical cupule do have a xylem- 
phloem orientation opposite to that of the larger bundles 
within the rachis, if one follows inward along the series 
of bundles in such wings, the orientation of each bundle 
is found to twist gradually, so that the innermost ones 
have the same orientation as the larger bundles (Plate 
III, fig. 1). This twisting of bundles suggests that the 
cupule wings are formed in part by a gradual folding out 
of flaps dislocated from the rind of the rachis. Further 
evidence in support of this view comes from the two- 
ranked spikes of our teosinte-derivative of Guarany 
maize. ‘The wide spacing of the cupules in this stock re- 
veals that the rind from the barren rachis has exactly the 
same vascular pattern as that in the cupule wings, and 
that the tissue at the back of the cupule is devoid of 
these small bundles (Plate IV, fig. 2). It seems, there- 
fore, that the portion of the cupule wings which includes 
the vascularization is derived from the rind of the rachis. 
[ 10 ] 
