﻿62 cox. 



It will be noticed in the above curves that the differences a and b 

 represent the sum of two factors, the change in results due to the 

 different heat treatment and the deviation due to the presence of loosely 

 held water; c and d represent only one factor, the deviation due to the 

 presence of water. The alteration of a single condition in the method 

 used in the analysis of coals to which the official method is applicable, 

 as represented by c and d produces more variation than the combined 

 changes in the method of the analysis of coals which give large mechanical 

 losses by the official method. I have already shown that the effect of 

 water when the official method is used is to increase considerably the 

 percentage of volatile combustible matter in Negros coals ; hence it must 

 be concluded that the factor represented by a or b minus this increase — 

 that is, the change in results due to the difference in the breaking down 

 of the hydrocarbons of the coal under the varied heat treatments of the 

 official and the smoking-off methods — is reasonably small. 



On the other hand, direct data have been given to show that there 

 is a positive difference. My results give discrepancies between the fixed 

 carbon and ash as determined by the two methods which are not direct!}' 

 proportional and therefore are not due wholly to mechanical losses. 

 Fortunately, this difference due to the varied heat treatments of the two 

 methods is smallest in the case of coals which can only be inaccurately 

 analyzed by the official method; it is also in these coals where large 

 variations in the percentage of moisture are apt to take place. Moisture 

 has no influence on the fuel ratio calculated from results obtained by 

 the smoking-off method, while large variations by the official method in 

 case it were applicable — that is, barring mechanical losses — would be 

 anticipated. Therefore, the variation produced by the altered breaking 

 down of the coal by the different heat treatments in the two methods 

 is probably not greater and perhaps even less than that which would be 

 possible with the official method alone; 17 while with the smoking-off 

 method the mechanical losses are overcome. 



In describing the coke ovens of the Colorado Fuel and Iron Company 

 at Segunda, Colorado, Mr. Hosea says: 18 "The larry cars hold 5J to 6 

 tons of disintegrated washed coal, and each weighs 12 tons loaded. This 

 is the customary charge for ovens of this pattern, and the charge is 

 coked in forty-eight hours, producing from 3^ to 3§ tons of coke, or a 

 yield of from 60 to 65 per cent." It is evident from the above that the 

 coal in the furnace is more slowly heated than it is by the official method, 

 although the latter is intended to approach the actual conditions in 



17 The results of Table 10 show that the greatest change in the percentage of 

 fixed carbon is produced by a variation of the loosely held water by not more 

 than 5 per cent. Unless care is exercised such variation may occur while the 

 sample is being taken and during its transportation to the laboratory. Cf. Som- 

 ermeier, E. E.: J. Am., Chem. Soc. (1906), 28, 1630. 



1S Hosea, R. M. : Mimes and Minerals Denver, Colorado (1904), 25, 8. 



