30 ROYAL SOCIETY OF CANADA 
followed that of the peats, lignites and bituminous coals and conse- 
quently the history of a coal can be placed in diagrammatic form by a 
compilation of the analyses of the different grades. 
LAOS Po 
BITUMINOUS COAL 
ANTHRACITE 

G RAPH/TE 
In this way we find that for each stage in the development we have 
a fairly constant decrease in oxygen with the increase in carbon. Also 
we have a mean hydrogen content which shows a slight decrease com- 
pared with that for oxygen. The variations for hydrogen above or 
below the mean as previously mentioned accompany certain characters 
in the coal. 
If we now attempt to follow the alterations by suppositions as to 
the materials lost to the mass we may first assume that the loss was 
possibly in hydrocarbons. Those that are indicated in the application 
of heat are the loss of carbondioxide and marsh gas. The relation of 
these, in order that the material left conform in analysis to that of the 
diagram, should be near the following ratio by weight és =$. At the 
same time we must allow for the reduction of the nitrogen by the for- 
mation of ammonia. Under this supposition, when the coal reached 
the anthracitic stage, the possible loss of carbon would be as high as 
sixty per cent of the original carbon or thirty per cent of the original 
mass. The carbon remaining (practically the whole mass) would be 
