76 IMPEOVING THE QUALITY OF WHEAT. 



not siifRcient to counteract the increase in nitrogen, and the result 

 is to increase the production of proteids per acre. 



Viewed in the hght of these various experiments, the selection of 

 large, heavy wheat kernels for seed does not appear to be altogether 

 unobjectionable, as in this case it resulted in a decreased production of 

 proteids per acre, without a compensating increase in the yield of grain, 

 when contmued for a numbei of years. On the other hand, the selec- 

 tion of the small, light seed is hardly to be recommended. In fact, 

 selection based upon kernel size or weight is not a satisfactory method 

 for permanently improving wheat. The individual plant should be 

 taken as the basis for selection, and very large numbers should be 

 handled. The figures in Table 8 show what great opportunity there 

 is for securing not only kernels of high nitrogen content, but also 

 plants giving at the same time an increased yield of grain and abun- 

 dant production of proteids. If the average nitrogen content and 

 yield of grain b}" plants be observed in this table, it will be seen 

 that numerous plants may be selected that have not onty a nitrogen 

 content above the average, but also a greater yield of grain. While, 

 therefore, it is probable that improvement in yield of grain can not 

 be effected so rapidly where it is combined with improvement in 

 nitrogen content as if the latter were neglected, yet present yields 

 of wheat in Nebraska can be increased at the same time that the 

 production of proteids is augmented. 



METHOD FOR SELECTION TO INCREASE THE QUANTITY OF 

 PROTEIDS IN THE KERNEL. 



The following tables show the results of analyses of a total of 

 forty-eight spikes of wheat. In the case of each spike one row of 

 spikelets, for instance, row No. 1, was analyzed, and the other row 

 of spikelets, which would then be row No. 2, was analj'^zed sepa- 

 rately. In the case of the set of spikes forming Table 14 the total 

 organic nitrogen was determined in both lots, and in the set com- 

 prised by Table 15 the proteid nitrogen was determined. The last 

 column shows the difference between the nitrogen content of the two 

 rows of kernels. 



