8 



Journal of Agricultural Research voi. xvin, No. i 



its absolute amount remained nearly constant. The dry matter and the 

 nitrogen-free extract again parallelled each other, while the protein, ash, 

 and ether extract remained rather constant. 



Figures 7 and 8 contain graphs for the composition of the bagasse. 

 Great reliance can not be placed upon these data, since the amount of 

 juice obtained from the cane varied from 34.5 to 37 per cent of the weight 

 of the cane. The bagasse as analyzed thus contained a considerable 

 proportion of juice. The relative proportions of the various constit- 

 uents are similar to those of the leaves and tops. 



Some interesting relations can be seen by collecting the percentage 

 curves for individual constituents for various parts of the plant on the 



Fig. 



NO. DAYS FROM DATE ?ANICLES APPEARED 



-Development of the proximate and mineral constituents of the seed heads of sorghum. 



same chart. In this way the history of each constituent can be viewed 

 in its relation to the whole plant. 



Figure 9 contains the dry matter curves. As would be expected, 

 there is a general increase in the percentage of dry matter throughout 

 the whole plant. The most marked increase is in the tops. The apparent 

 abnormality of a sharp decrease in the last period of this curve is unex. 

 plained; it is no doubt an analytical error. A large portion of the 

 increase in dry matter of the bagasse is probably due to the sugars of the 

 retained juice, since the two curves parallel each other very closely, and 

 since only about two-fifths of the juice was expressed by the small ex- 

 perimental mill. 



Figure 10 contains the curves for the crude protein. As has been 

 pointed out above, there is not only a regular decline in the percentage 



