90 THE ROYAL SOCIETY OF CANADA 



of the original coal, the deduced loss of volatile matter on the moisture 

 free coal basis, and the calculated analysis of the carbonized residue. 

 The calorific values are taken from the curve for the regular series of 

 tests. 



The residue from a moisture determination, in an atmosphere of 

 carbon dioxide in a toluol oven, is taken as the moisture-free basis 

 for all calculations and curves, it being assumed that no volatile 

 matter is driven off at 110°C. In the calculations for Table III, the 

 percentage of volatile matter found by the standard method of deter- 

 mination is assumed to be the total volatile matter in the coal. The 

 regular series curve in Diagram I indicates that this is approximately 

 correct. It is also assumed that if a certain fraction of the volatile 

 matter in a sample is expelled during an experiment, the residue 

 will contain the remainder as volatile matter. This is doubtless 

 approximately correct, although decompositions may take place 

 whereby the residue on analysis will show less than the calculated 

 content. 



Regular series of tests: — Nine completed tests were run in this 

 series over a temperature range of from 200°C., to 700°C., the general 

 procedure given above being followed. In addition, moisture deter- 

 minations in the toluol oven were regarded as giving the comparable 

 results for a temperature of 110°C., and standard volatile matter 

 determinations, in a platinum crucible over a gas fîame, the results 

 for a temperature of approximately 900°C. 



The results obtained, as shown in the curves, indicate that the 

 loss of volatile matter is slight up to 300°C., from there up to 600°C. 

 the loss increases rapidly, and by 700°C. the volatile matter has nearly 

 all been driven off. The calorific value of the residue increases steadily 

 up to 550°-600°C., and then commences to decrease. 



A very striking phenomenon, first observed in connection with 

 the vacuum series, was later found to take place with every sample of 

 dried or carbonized lignite. The residue in every case rapidly gained 

 in weight after removal from the retort, even when stored in a desic- 

 cator over sulphuric acid ; its calorific value at the same time decreasing. 

 The results given in the curves and in the tables are, with the exceptions 

 noted, for weights and calorific values determined immediately after 

 the experiment. When the tests described in this report were first 

 made, the calorimetric determinations were not carried out immedi- 

 ately, the delay in some cases being as much as two or three days. 

 The loss in calorific value on standing, however, was so serious that 

 it was found necessary to repeat all the more important experiments 

 with the time factor standardized as already described. As an 

 example of the above: — one sample, carbonized at 550°C in the 



