SELECTION TO INCREASE PROTEIDS IN KERNEL. 



77 



TABLE 14. Analyses of twenty-jive spike* of wheat, showing their total organic nitrogen. 



TABLE 15. Analyses of twenty-three spikes of wheat, showing their percentage of proteid 



nitrogen. 



Percentage of proteid 



Percentage of proteid 



It will readily be seen that the analyses of the rows agree very 

 closely, the extreme difference being 0.22 per cent, and the average 

 difference being 0.12 per cent, in the total nitrogen. If, therefore, 

 one row of spikelets were to be used for seed and the other were 

 analyzed, it is quite evident that a very accurate estimate of the 

 nitrogen content of the kernels used for seed would be obtained. In 

 the determination of proteid nitrogen there is an extreme 'difference 

 of 0.36 per cent in one case, but in the main the differences are small. 

 As will be shown later, the variation in the proteid nitrogen content 

 of individual plants is so great that even this maximum difference 

 would cause no confusion when selecting plants for reproduction. 



It is very desirable to have for analysis a larger .sample than can 

 be obtained from one spike. It has therefore been attempted to 

 ascertain whether a sample consisting of one-half the whole number 

 of spikes on a plant would afford a fair estimate of the composition 

 of the other kernels on the remainder of the spikes. The plants 

 whose spikes were analyzed were grown in hills 5 inches apart 



