— 24 — 



F. — Similar to E, temperature iio°. 



G. — Similar to D, dry hydrogen being used in place of air, tem- 

 perature 97°. 



H. — Similar to G, temperature ioo°. 



I. — Similar to G, temperature iio°. 



J. — Drying in a vacuum, over sulphuric acid for 72 hours. 



Three gms. of substance were used in every case, and the 

 samples were cooled in a desiccator thirty minutes ]:>efore weighing. 

 Complete drying was assured by further treatment for one hour, 

 and re-weighing ; except when otherwise mentioned the full time 

 of treatment was three hours. 



The results obtained are shown in the following table, each re- 

 sult being the average of four determinations, the highest and 

 lowest agreeing within .15 per cent. 



The conclusions to be arrived at from inspection of the foregoing 

 table are — 



1. Raising the temperature raises the percentage of loss, as 

 would be expected. 



2. The loss at 100° in an air bath corresponds clo.sely to the loss 

 in a vacuum. 



3. The effect of the temperature of 110° is greater in the deter- 

 mination on the hay, than on the bran and cotton seed meal. 

 This seems to indicate either the presence of a greater percentage 

 of easily volatilized constituents in the hay, or that the hay gives 

 up its moisture with more difficulty. 



The first explanation is supported by the results of the following 

 experiment. The current of air after passing through the sub- 

 stance [methods D. E. F.jwas passed over red hot copper oxide, 

 through a series of drying tubes containing sulphuric acid, and 

 finally through a Geissler potash bulb. The gain in weight of 

 this bulb, due to the carbonic acid formed by the oxidation of the 

 volatile carbonaceous constituents carried along by the current of 



