were able to obtain. The majority of the cobs from the three 
lowest levels were eight-rowed and none had more than ten 
rows. In contrast the majority of cobs in the two upper levels 
had more than eight rows and several had twelve or fourteen 
rows. One exceptional cob in level 3 (not included in Table 1) 
was distichous and had only four rows. 
Accompanying an increase in kernel-row number was an 
increase in the diameter of the rachis and the total number of 
spikelets. 
This conclusion is more a matter of observation than one of 
statistical averages since the number of intact cobs was limited. 
A trend can be illustrated, however, by comparing the intact 
cobs from levels 2, 3 and 4. The averages from these three 
levels are 133, 126 and 113 spikelets respectively. 
The early cobs from this site appear to represent a weak form 
of pod corn. The glumes are quite long in relation to the diame- 
ter of the rachis and they are soft and fleshy; not like the 
indurated lower glumes of corn’s relatives, teosinte or 7rip- 
sacum. 
With respect to known races of maize of Peru, some of these 
cobs could be assigned to the Peruvian popcorn race Confite 
Morocho described by Grobmanet al. (1961, Fig. 49). Like the 
slender cob illustrated by these authors the rachis is square in 
cross section and the cupules are shallow in outline. None of 
the cupules are as long as those of the unusually slender ear 
illustrated in Grobman et al. Cobs of this type are also quite 
fragile, breaking up easily into sections of one cupule each. 
Some of the smaller cobs have the stumps of terminal staminate 
spikes. 
The early cobs, which comprise the majority of the collec- 
tion may be related to the Peruvian race, Confite Morocho. 
This possibility is supported by cob morphology, number of 
kernel rows, which approaches an average of 8, kernel type, 
form and consistency of the glumes, etc. They bear also some 
resemblance to the Mexican race Chapalote, although they 
seem to be farther removed from it. Except for one cob, with 
four rows of spikelets and stiff lower glumes, which we would 
have classified as tripsacoid had we found it in the context of a 
Mexican ancient cob collection, all the others exhibited long, 
Zed 
