pressed. The genotype of modern corn may well be a constella- 
tion on a grand scale of DNA recombinants. 
So far little progress in corn breeding has been made by 
hybridizing corn directly with teosinte but phenomonal results 
have been obtained by combining strains that are the product of 
past introgression from teosinte as shown by anatomical and 
genetic studies (32). 
The average yields of corn in the United States have in- 
creased from 26 bushels per acre in 1929 to 95 bushels in 1972. 
This progress has involved the bringing together by empirical 
methods of strains of germplasm from diverse sources, much of 
it originally from teosinte. There are many races of corn not yet 
employed in hybrid corn breeding (33). Recognizing the poten- 
tial value of these for breeding opens new possibilities for 
improvement that may render corn, this nation’s basic food 
plant, still more important on the world scene as one of not 
more than about twelve species of cultivated plants, each one a 
unique biological system, that quite literally stand between 
mankind and starvation. 
SUMMARY 
Fossil pollen discovered in core samples from a depth of 69.3 
- 70.5 meters in preparation for the construction of Mexico's 
first skyscraper has been identified as the pollen of a primitive 
wild corn. Its authenticity seems now to be well established 
since the possibility that the pollen represents modern con- 
tamination is shown to be remote. The authors conclude that 
the ancestor of cultivated corn was a wild corn and not its 
closest relative, teosinte, and that this may have important 
implications for corn’s genetic improvement and its role in 
meeting the world food problem. 
POSTSCRIPT 
Two important events occurred while this paper was in pro- 
of. Anarticle by Iltis et a/. (appearing in Science, Jan. 12, 1979) 
251 
