370 C. R. HENDERSON 



Although there is a considerable body of literature on heritability esti- 

 mates, we need more accurate estimates of the heritabilities of most traits of 

 economic importance. Also almost nothing is now known about genetic cor- 

 relations between traits, about genetic-environmental interactions, and 

 about the magnitude of genetic differences among herds. Estimates of these 

 genetic parameters are essential to intelligent selection for additive genetic 

 values. In the case of inbred lines, little is known concerning the variances 

 of general and specific combining abilities. The work of Sprague and Tatum 

 (1942) with corn and Henderson (1949) with swine illustrates the types of 

 estimates which are badly needed in selecting for general and specific combin- 

 ing abilities from the results of line cross tests. 



Finally, well designed experiments are needed to test how closely predic- 

 tions made from indexes or other selection procedures check with actual re- 

 sults. 



