176 Mr. Dalton on the constitution of the atmosphere. 
All our ordinary experiments are limited to given volumes 
of gases which are considered of equal densities and tempe- 
ratures throughout the volumes ; but this is not the case 
when the volume is large, and extended in a direction per- 
pendicular to the horizon ; such, for instance, as a cylindrical 
column of the atmosphere of the altitude of several miles. 
The density in such vertical columns diminishes nearly' in 
geometrical progression to equal intervals of ascent, and the 
temperature diminishes nearly in arithmetical progression, 
or in direct proportion to the ascent. 
When we observe the diffusion of two gases (A) and (B) 
through each other in small limited volumes, such as in our 
ordinary experiments, we may ascribe it on the one hand to 
the mutual attraction of the particles of (A) and (B); or on 
the other, to the repulsion of the particles of ( A) for each 
other, and their non-repulsion of those of (B); and vice versa. 
The effect would seem to be precisely the same on both 
views. But it is not so when we consider a vertical atmo- 
spherical column of mixed gases. Here the two views display 
their difference in a striking manner. 
An exposition of the results of two indefinitely long per- 
pendicular columns of any given gases (A) and (B) being 
mixed together, having never yet been laid before the philo- 
sophical public according to either opinion, I apprehend it 
may be of service to the advancement of knowledge on this 
interesting subject to draw attention to the following con- 
siderations. 
Let A and B be two equal cylindrical tubes placed in 
contact and perpendicular to the horizon, of indefinite length, 
closed at the bottom and open at the top. Let the tube A have 
