170 Dr Fyfe on the Comparative Value of Oil and Coal Gas . 
into the retort At Leith, a gallon of whale-oil affords from 98 
to 108 cubic feet; and the same quantity of palm-oil, from 97 
to 114. It may be considered a fair estimate to obtain 100 feet 
from each gallon, presuming, of course, that the oil is decom- 
posed under the most favourable circumstances, so as to get a 
gas possessing the greatest illuminating power, for on this every 
thing depends. From experiments I have performed on a small 
scale, and from trials made at Leith, I find that if the oil be al- 
lowed to flow into a retort brought just to a red heat, there is 
comparatively little gas, but a great deal of volatile oil. When 
the retort is brought to an intense heat, lamp-black is formed in 
considerable quantity ; so that in both of these ways there is a 
great loss. When the retort is at a full red heat, the oil seems 
to undergo decomposition most easily, and to give off the largest 
proportion of good gas. 
Illuminating Power of Oil and of Coal Gas. 
Various statements have been given of the illuminating power 
of oil and coal gas ; nor is this to be wondered at, when we consider 
that the quality of the gases depends so much on the mode of 
preparation, and take into account also the defective modes usu- 
ally adopted for determining the intensity of the light afforded 
bv their combustion. Mr Ricardo, in his early papers on this 
subject, has given a very flattering account of oil-gas. Fie states, 
that an Argand burner, giving a light equal to six candles, six 
to a pound, consumed one cubic foot per hour : and as Mr 
Accum mentions, that an Argand of coal-gas, giving a light 
equal to three candles, eight to the pound, consumes two feet 
in the same time, he has inferred, supposing the candles of the 
same size, the illuminating power as 4 to 1 ; but taking the 
average of a number of experiments, he has fixed their compa- 
rative power of giving light as 3 J to 1. 
In these trials, however, the gases were not brought into com- 
parison with each other by burning them together, and the data 
on which he proceeds seem to be very fallacious, as it is not 
stated whether the candles were of the same kind in both ex- 
periments. Messrs Taylor and Martineau, the patentees of the 
oil-gas apparatus, have, however, come to nearly the same con- 
clusion, that the illuminating power is as 3 J to 1, a conclusion 
drawn from the experiments of Mr Brande and Mr Faraday. 
