maize which would be one parent of the original hybrid 
with Manisuris was probably a form of pod corn and since 
the tunicate locus which is responsible for pod corn is 
dominant over the non-tunicate condition in hybrids of 
tunicate maize with both teosinte (Galinat, 1959) and 
Tripsacum (Galinat, unpub. ), it seems probable that the 
original Tripsacum would have had the slender rachis 
with small cupules and the long herbaceous glumes char- 
acteristic of pod corn. Sometime during its evolutionary 
history, these may have been replaced by the thick rachis 
with large cupules and the indurated outer glumes char- 
acteristic of the non-tunicate condition and of modern 
Tripsacum. 
Despite its slender rachis segments, Manisuris has a 
substitute for a cupule or a false cupule formed by the 
close position of sterile pedicels along alternate edges of 
successive rachis segments. Possessing this pedicel- 
formed cavity, it required only the true cupule from 
maize to form the cupulate fruit case characteristic of 
Tripsacum. Essentially unchanged from Manisuris 
might have come the sessile spikelets borne parallel to 
the rachis in a manner in which they could develop within 
the protective confines of a true cupule once it was intro- 
duced by maize germplasm. Also derived from Manisuris 
are the distichous arrangement of the spike and, at ma- 
Tasie II. Pollen size variation in microns? 
Grass pd Min. Max. o n 
Manisuris cylindrica 37.4 28 46 4.10 93 
Tripsacum” 40,4 26 58 5.28 489 
Teosinte ® 73.0 56 86 5.10 233 
Maize (Chapalote) S782 74 102 6.12 228 
‘From stored pollen that has been swollen with 85% lactic acid 
and stained with IKI solution. 
? Pooled data from T\. zopilotense, T. floridanum and T’. dactyloides. 
* Pooled data from Florida teosinte and Arcelia teosinte. 
[ 297 ] 
