104 



recorded in Bulletin 17 and Professor Sbubbs's Bulletin Ho. 12, it ought 

 to be possible to weigh them justly and to form some approximately 

 accurate idea of the value of sorghum as a sugar producer. 



First of all, it will be necessary to divide the analytical data into two 

 classes, viz: (1) Data derived from tha analyses of small samples, in 

 other words, experimental data, and (2) those obtained from analysis 

 of large quantities of material entering in the process of manufacture ; 

 in other words, manufacturing data. 



I have collected below all the mean analyses of experimental samples, 

 and have obtained therefrom a general average of the character of all 

 the juices which have been analyzed in a small way. 



BY THE DEPARTMENT. 



