406 Experiments on the Gas from C oal, 
gas, which were too minute to be wilde ti 
‘measured. 
In the small experiments made several 
years ago, I never found, in the early pro- 
ducts of gas from cannel coal, a proportion 
of olefiant gas at all approaching that which 
is noted in Table I, and its quantity in small - 
distillations rapidly decreased, until in the 
latter products it could be no longer traced 
at all. ‘The method of analysis, which I for- 
merly employed, led me, however, as I have 
lately discovered, to under-rate the propor- 
tion of olefiant gas, and to over-estimate that 
of sulphuretted hydrogen. But making due 
allowance for this error, the superiority of 
the products of large operations, so far as 
respects olefiant gas, still exists, and is con- 
firmed by comparative experiments on a small 
scale which 1 have latelymade. Thus it ap- 
pears from table I, that even after twelve 
hours continuance of the process, olefiant 
gas still constitutes 4 per cent. of the gases 
evolved from cannel. The other inflammable 
gases, also, when obtained in large quantity, 
are more uniform in quality, and possess, to- 
wards the close of the process, much greater 
combustibility and illuminating power, than 
when procured in small experiments. This 
superiority is obviously dependent on the 
