176 



Huh tt, Mack and ISmylh — 



ated along with the moisture and that they ceased to 

 conic off when the moisture did, so it seemed probable 

 that these gases (mostly C0 2 with some CO and a little 

 CH 4 ), were also absorbed by the large surface of the col- 

 loidal coal in much the same way that the water was. 

 We arranged our apparatus so that we could follow the 

 liberation of these gases manometrically, and were thus 

 able to plot the volume of gas liberated as a function of 



Fig. 1. 



IS - 



°" ffi 

 x // / 











HO 











?■ l\i 













^^ 18t- 



It. 









lSt'io^n) 





«-< // / 













1+. 



jJ // / 











130- C ' 



1.2- 















10- 























105' 1-) 



0.8 • 















O.b- 







P.tUb 



jrgk 



Cofxl 



0.4- 















02- 





M 



iT\utes 









20 



U 



the time of heating for each temperature. These iso- 

 thermals were very similar to those previously obtained 

 for the moisture losses as a function of time, and were 

 much easier to determine as it only required manometer 

 readings at convenient time intervals, while in the pre- 

 vious work it was necessary to make a separate deter- 

 mination of the moisture for each point on the curve. 

 In this work we were also able to make a determination 

 of the total water liberated at a given temperature, while 

 the plot of the gas liberation curve showed whether there 

 had been noticeable decomposition. 



An idea of the apparatus and method may be obtained 

 from figs. 2, 3 and 4. The coal bucket was made from 



