EARLY TESTING AND RECURRENT SELECTION 



407 



have presented data on a comparison of combining ability in F2 and F3 lines 

 of corn. On the basis of these comparisons they concluded that early testing 

 was of doubtful practical value. The material used in this study was 30 

 selfed ears from early segregates from the single cross A116XL317. Each of 

 the 30 selfed ears was grown ear-to-row and pollen from approximately 30 



IK) 



105 



100- 



THREE HIGH COMBINING S, LINES (MEAN) 



-THREE LOW COMBINING S, LINES (MEAN) 



\ 



\ 



60 



75 



\ 



^Jc— 



r 



GENERATIONS of INBREEDING 



Fig. 26.2 — The effects of visual selection and testing for coml)ining ability during four 



generations of selling in the variety Krug. 



plants in each progeny was bulked and applied to the four inbreds chosen as 

 testers; A334, A357, A340, and A392. 



In addition, five individual plants selected at random were also out- 

 crossed to the same four testers. The test crosses arising from the bulked 

 pollinations were considered as representing a random sample of the gametic 

 production of the individual Fo plants and the five individual test crosses as 

 samples of the F3 progenies. Adequate seed was obtained from 26 of the origi- 

 nal 30 families. Within the different tester groups correlations between F2 

 and F3 test cross means ranged from .51 to .76. 



