EARLY TESTING AND RECURRENT SELECTION 



405 



ance were no more heterozygous than the So plants having poor test cross per- 

 formance. The distributions arising from the four highest yielding families 

 were not significantly different, but were significantly different from the 

 distributions arising from the two lowest yielding families. These same general 

 types of results were obtained when stalk breaking was considered. 



Finally, three of the lines arising from selected sample when in the S3 gen- 

 eration of inbreeding were compared with five standard lines. These eight 

 lines were crossed in all possible combinations and compared in rej)licatcd 

 yield trials (Table 26.2). The S3 lines, as a group, were superior to the stand- 



TABLE 26.1 



FREQUENCY DISTRIBUTION OF ACRE YIELDS IN BUSHELS FOR 



20 Si TOPCROSSED PROGENIES DERIVED FROM 



6 So LINES (SPRAGUE, 1946) 



TABLE 26.2 



RELATIVE AVERAGE PERFORMANCE OF 

 STANDARD AND NEW S3 INBRED LINES 

 OF CORN BASED ON SINGLE CROSS YIELD 

 TRIALS (SPRAGUE, 1946) 



