28 Drs Christison and Turner on the Construction of 
tive proportion of the extraneous gases to each other, some being more inju- 
rious than others ; so that an increase in the quantity of the illuminating in- 
gredients would not only add to the light of that additional quantity, but like- 
wise diminish the loss occasioned by the contaminating gases. 
These considerations led us to doubt strongly the accuracy of Dr Fyfe’s 
proposed method. But we have said that he found it to correspond in its re- 
sults with the inferences drawn from actual measurement of the light. For rea- 
sons formerly mentioned, however, his measurements appear to present intrin- 
sic evidence of their inaccuracy ; and certainly our own trials, though they 
have once or twice agreed with his, do not by any means show the uniform 
correspondence for which he contends. On one occasion, when the loss by the 
action of chlorine was 14 and 34, the illuminating power was as 100 to 233, 
instead of 243. Here the agreement is tolerably close. But, on another oc- 
casion, when the relative condensation by chlorine was 16 and 46, the illumi- 
nating power was only as 100 to 250, instead of 287 ; and again, when the re- 
lative loss was 13 and 37, the relative illuminating power was only 100 to 225, 
instead of 284. 
There is still a third method, which was proposed by Dr Henry himself, 
but which is even less accurate than the two foregoing, and has since been 
abandoned, we believe, by its ingenious author. It consists in detonating the 
gas with oxygen, and valuing its power of illumination by the relative quan- 
tity of oxygen that disappears. Several objections might be stated against 
this plan ; but it will be sufficient to mention one only. As in the case of the 
specific gravity, a certain ratio may exist between the illuminating power and 
the quantity of oxygen which disappears. But the ratio has not been deter- 
mined by actual measurement ; and, at all events, it is not, as Dr Henry at 
first supposed, a simple one. For, according to his own tables, the relative 
light of two gases having the specific gravities of 620 and 906, is only 3 to 4; 
while we have found that Count Rumford’s photometer indicates a proportion 
of 1 to 2. 
At present, therefore, no method is known for ascertaining the relative 
light of the gases, except by actual measurement of it. A plan, indeed, 
has occurred to ourselves, which will have the peculiar advantage over 
every other, of giving accurate indications even in the hands of a com- 
mon workman, and, consequently, of being fit for the every-day use of Gas 
Companies. But we have not yet finished the observations on which the 
scale must be constructed. It consists in simply burning a jet of a certain 
length through an aperture of a given diameter, from a small gasometer of 
perfectly steady pressure, and noting the expenditure. So far as we have 
hitherto tried, the expenditure will be a very accurate criterion of the light. 
But the ratio is not a simply inverse one. Thus, when a 4-inch jet requires 
for its support twice as much of one gas as of another, the proportional light 
is not as 1 to 2, but as 1 to 2| fully. In fact, the flames, though equal in size, 
are very different in brilliancy. As the exact ratio, however, has not been 
determined by an adequate variety of experiments, we shall confine our sub- 
sequent statements to the results' drawn from direct photometrical measure- 
ments. 
