EXPERIMENTS IN BREEDING SWEET CORN. 



271 



30 



























25 











j 

















ti 20 



























a 15 















\ 













10 









y 



1 





\ 













5 





/ 





/ 

























/ 

















■^ 







21-24 24-27 27-30 30 33 33-36 36-39 39-42 42-45 45-48 48-51 51-54 

 BUSHELS PER ACRE 



Fig. 223. Diagram showing the yield of the lOO rows of Type I corn 

 in bushels per acre. 



From this diagram it appears that the yield which occurred 

 most often was in the class 33-36 bushels. From this high 

 point the polygon slopes off at about the same angle in both 

 directions — i. e., towards higher as well as lower yields. 



5. Even the 11 best rows of the Type II corn, selected be- 

 cause they were the best in the field, did not average to yield 

 as well as the whole of the 100 rows of the Type I, good, bad 

 and indifferent taken together. The Type II falls behind nearly 

 6 bushels to the acre. 



6. Further, even in these 11 best rows the Type II corn did 

 not produce nearly so high a proportion of its total shelled corn 

 on ears of good quality which warranted their use for seed. 

 Thus, taking averages of the 1 1 rows it appears that of the total 

 yield 48.2 per cent, was borne on nubbins or ears not good 

 enough to use for seed. This figure is to be compared with the 

 15.2 per cent, in the case of the Type I corn. 



7. The shelling ratio is also slightly higher in the Type I 

 averages for the whole field than in the case of the 11 selected 

 rows of Type II. From Table 3 it appears that in the case of the 

 "Ai seed" of Type I, 17.5 per cent, of the average weight of the 

 ear was cob (shelling perccntage=82.5), and in the "good seed" 

 of the same type of corn, 18.9 per cent, of the average weight of 

 the ear was cob (shelling pcrccntage=8i.i). On the other 



