490 



JOHN W. GOWEN 



son, and even between simultaneous carefully controlled experiments, it is 

 not surprising that the general and specific inheritance effects show varia- 

 tions. A particular fitting of strains to place and season, etc., appears essen- 

 tial for highest yields. In view of this conclusion, it is important to remember 

 that this effect is within hybrids, and that hybrids, in general, are distinctly 

 better than the inbreds (see Fig. 29.3). 



Through the kindness of Dr. G. F. Sprague, making available certain of 

 his extensive data on Fi crosses of some 62 inbred lines of corn, we have been 

 able to extend this analysis and compare the stability of general and specific 



O 



z> 

 o 

 o 



CL 



o 



< 



UJ 



140 



20 



100 



80 



INBRED X 

 SYNTHETIC y 



INBRED LINE 

 CROSSES 



■•••s.x 



INBRED LINES 



15 



20 



25 



30 



35 



GENERATIONS OF BROTHERxSlSTE R MATING 



Fig. 29.3 — Changes in productivity with advancing generations of brother X sister mat- 

 ing of inbred Unes, inbred line crosses, and inbred X s_vnthetic. 



combining ability in the two species. The trials were conducted yearly from 

 1940 to 1948. The Fi hybrids were planted in ten different areas chosen as 

 representative of the different climatic conditions of Iowa. Any one trial may 

 contain all possible single crosses of 4 to 14 inbred lines. The trials contain 

 many individuals, and are replicated several times so the record for the Fi is 

 an average of numerous Fi's instead of an individual as in the data on Dro- 

 sophila. As would be expected, a particular cross was occasionally lost from a 

 test. When this happened, the missing plot value was calculated from the 

 mean and the general combining abilities of the lines entering the test. The 

 specific value was considered zero. The data for the general and specific com- 

 bining abilities of the lines in the remaining plots were used for further study. 

 Our study considered the first order values, as these are the only values which 

 have operational significance in breeding for heterosis. 



