Moisture Content of Some Typical Coals. 175 



On increasing the vapor pressure the coal took up water 

 again, but when the water removed had been returned, 

 the vapor pressure was not that originally shown by the 

 coal but measurably greater. This is a typical behavior 

 of colloids. 



The water content of a coal must, therefore, be denned 

 as the moisture that is present as such when the coal is 

 first broken down from the seam, and care must be 

 exercised to avoid loss or gain of moisture during the 

 sampling, collecting, or handling of the samples. 



The present work was on four samples typical of 

 American bituminous coals. These samples were secured 

 by engineers of the U. S. Bureau of Mines, reduced to 60- 

 mesh samples and given the regular analysis at the Pitts- 

 burgh Station: 





New Eiver 



Pittsburgh 



Indiana 



Wyoming 



Moisture 



110 



1-93 



4-55 



6-15 



Volatile 



24-67 



36-23 



41-65 



37-55 



Fixed Carbon 



69-98 



55-42 



42-70 



52-30 



Ash 



4-25 



6-42 



11-10 



4-00 



Precautions were taken to avoid a loss or gain of 

 moisture after the 60-mesh samples were prepared, but 

 these samples would not comply with the criterion given 

 above, as we do not as yet know the loss or gain of mois- 

 ture in preparing the 60-mesh samples. The question of 

 local heating due to crushing may play a measurable role 

 and must be looked into, and until information on these 

 points is available we can not consider the values we have 

 obtained as representing the moisture content of the coal 

 as mined but only of the 60-mesh samples. 



From our previous experiences we were able to materi- 

 ally improve our method and apparatus. The coal was 

 heated in a vacuum so devised that the water was 

 removed from the presence of the coal as fast as liber- 

 ated. It was found that for a definite constant temper- 

 ature the moisture came off rapidly at first, then more 

 slowly and finally ceased altogether, thus giving results 

 which were plotted as "percentage loss of water — time 

 curves." These curves showed a rapid evolution of 

 water at first but the curves finally ran practically paral- 

 lel to the time axis, showing no further liberation of 

 water. Fig. 1 represents the curves obtained at the 

 various indicated temperatures for a Pittsburgh coal. 

 We noticed in the previous work that gases were liber- 



