lished) has identified prehistoric Tripsacum, apparently 
T. dactyloides, from FE] Diablo Cave, Tamaulipas, Mex- 
ico and has found the same form growing in the vicinity. 
A general appraisal of the evidence published to date 
suggests that, of all forms tested, a 2n=36 type of 7. 
dactyloides is the one most likely to produce a fertile, 
intermediate, backcross hybrid; this has virtually all the 
plant characters that would be required. ‘The main ob- 
jection is that, in combination with corn varieties with 
which it has been tested, this type falls a little short of 
being sufficiently interfertile. In 1989, we (26) pointed 
out several additional characters which, at that time, 
seemed to make it unsuitable as a putative parent. Al- 
though most such characters now have been explained in 
accordance with the theory of hybrid origin (84), a form 
of Tripsacum which meets the requirements better than 
those now known may yet be found in the future. 
It seems desirable to consider here the implications of 
the suggestion made by Anderson (1), and now supported 
by Farquharson’s experimental results (13), that the 
form of 7'.dactyloides currently regarded as diploid might 
be in reality an allotetraploid. If the center of origin of 
Tripsacum is in Mexico or Guatemala, as both W eather- 
wax and Randolph conclude, it would be natural to sup- 
pose that either 7°. dactyloides (2n=36) or its parents 
once occurred there. In this event, one of the parents 
(2n=18) of 7’. dactyloides and a primitive corn might be 
the parents of teosinte (47). However, we do not wish 
to base the case for the hybrid origin of teosinte on an- 
cestors still unknown or now extinct. Rather we would 
emphasize that teosinte could be derived through the 
hybridization of existing forms. 
Tar Errectrs or TripsAcuM CHROMOSOMES AND GENES 
In our first hybrids between corn and ‘Tripsacum (26), 
[ 860 | 
