^36 



STATE BOARD OF AGRICULTURE. 



from variatiou in the size and weight of the berry, and not from any change in 

 the number of grains growing on a given area, since new grains would not 

 form during the period of ripening, and all those already formed would pass 

 througli the successive stages of development more or less completely. In this 

 way there were eliminated several sources of error incidental to measuring 

 the grain produced on a given area, such as the loss of grain in handling, and 

 from incomplete threshing, varying productiveness of different plots of soil, 

 etc. All broken kernels and foreign substances of every kind were removed, 

 but no grains were removed in consequence of imperfect development. The 

 grain was then poured into the scale-pan of a delicate balance, and ten 

 grammes carefully weighed out and the number of grains of wheat counted. 

 This was repeated ten times for each specimen and one-tenth of the sum of 

 the whole was taken as representing the number of kernels of wheat for the 

 weight of ten grammes. 



The gross product of grain at the several stages of growth would be directly 

 as the weight of the kernels, and inversely as the number of kernels required 

 to produce a given weight. 



The average number of grains of wheat required to weigh ten grammes at 

 the several periods of cutting is given in the following table. I have also 

 estimated the number of bushels of grain at the different periods of cutting, on 

 the supposition that each variety of wheat produced thirty bushels to the acre 

 as its maximum. The table also contains the percentage of nitrogen, cellulose, 

 and ash, at the different periods of growth : 



TABLE I. 



