548 
stands three or four inches higher within the tube than it does 
outside. Could we have observed what took place in the pores of 
the plaster of Paris plate, we should have seen the particles of 
coal-gas passing out of the tube and those of air passing into 
it, but the speed at which the coal-gas moved would be much 
greater than the speed of the air. The ultimate product 
within the tube is atmospheric air and water. If hydrogen 
be substituted for coal-gas in a similar experiment, the water 
will rise in the tube twice as fast and to twice the height 
than it did when coal-gas was used. 
Graham's law of diffusion may be shortly stated thus: 
" Gases diffuse into each other, or into space, in an inverse 
proportion to the square root of their densities/ 9 By this 
law a light gas diffuses rapidly, whilst a heavy gas diffuses 
but slowly. Coal-gas (light carburetted hydrogen, or fire- 
damp) represents a light gas, and carbonic acid (choke-damp, 
or after-damp) is a heavy gas, diffusing at only half the rate 
of common air. 
The influence of gaseous diffusion upon our atmosphere is 
all-important. By it the noxious gases resulting from the 
respiration of animals and decay are prevented from accumu- 
lating at any spot. 
"We now reach the second division of the subject, being 
the application of the foregoing. Mr. Geo. F. Ansell has 
devoted many years of anxious research to the adaptation of 
instruments capable of applying the laws of gaseous diffusion 
to the detection of fire-damp in coal mines, and his instru- 
ments represent an analogous position to the safety-lamp of 
Davy, since both have resulted from the logical application 
of a known law. 
A thin pellicle of india-rubber, though quite impervious 
to the passage of fluids, permits the free diffusion of gases, 
and Mr. AnselTs earliest instrument, which was made about 
four years since, consisted of a balloon of this material, 
