172 
* 
which is about 90100 feet, or between 17 and 18 miles, if 
convective equilibrium existed and if the gaseous laws had 
application to so low temperatures and densities. It has 
always appeared to the Author to be most improbable that 
there is any limit to our atmosphere ; and no one can suppose 
that there is a limit at any height nearly so small as 17 or 18 
miles. It is difficult to make even a plausible conjecture as 
to the effects of deviations from the gaseous laws in circum- 
stances of which we know so little as those of air at very low 
temperatures ; but it seems certain that the other hypothesis 
involved in the preceding equations is violated by actions 
tending to heat the air in the higher regions. For at 
moderate elevations above the surface, where we have 
air following very strictly the gaseous laws, the rate of 
decrease of temperature would, according to equation (4), be 
be 
*41 X T 
1-41 H 
per foot, that is to say, 
1 ° 
per foot, since 
T 
H = 26224 X or 1° cent per 329 feet. Now, the actual 
274 1 
decrease, according to Mr. Welsh, is 1° cent in 530 feet, or 
not much more than half that according to convective 
equilibrium. 
It seems that radiation, instead of partially accounting 
for the greater warmth of the air below, as commonly 
supposed, may actually diminish the cooling effect, in going 
up, which convection produces. In fact, since direct con- 
duction is certainly insensible, we have only convection 
and radiation to deal with, except when condensations of 
moisture, &c., have to be taken into account. In fair and 
cloudless weather, then, the lower and lowest air being on the 
whole warmer (the lowest being of course at the same tem- 
perature as the earth’s surface), it is perfectly certain that 
the upper air must gain heat by radiation from the lower — 
and that the convective difference of temperature must be 
diminished by the mutual interradiation. 
