Gradation of Heat in the Atmosphere. bt 
the greater or less degrees of moisture that prevail, are some of 
the causes that have most influence, more particularly on the tem- 
perature of the stratum of air in the immediate vicinity of the 
earth. But there is a predominant cause, which makes the tem- 
perature continually diminish as we ascend to greater heights in 
the atmosphere. We allude to that property of air, by which it 
absorbs heat when it expands by being less compressed. A por- 
tion of air that has become heated at the earth’s surface, rises up- 
ward on account of its diminished density; as it ascends, the 
pressure being less, it expands and becomes colder by absorbing 
heat ; and hence the velocity of its ascent decreases, and it finally 
comes to rest when its density is reduced to that of the surround- 
ing mass. ‘The property which air possesses of becoming colder 
by rarefaction, checks the elastic force by which it tends to fly 
off from the earth. It operates along with gravity to make that 
fluid cling to the earth, and to impose a limited boundary on the 
atmosphere, 
All our knowledge of the gradation of heat in the atmosphere, 
is derived from observations of the barometer and thermometer 
made for the purpose of measuring heights. ‘The exactness of such 
measurements, within certain limits, cannot be questioned; be- 
cause they have been verified in so many instances by comparing 
the like results obtained by levelling and by the operations of 
geometry. We may therefore presume, that the principles on 
which is founded the rule for calculating heights by the barome- 
ter, are nearly correct. Hence the great improvement of this 
method, introduced by De Luc, of estimating the temperature of 
the column of air at a mean between the temperatures observed 
at its extremities, must be at least a near approach to the truth. 
This estimation is equivalent to supposing that the heat decreases 
uniformly from the bottom to the top of the column ; which law, 
if it be not rigorously exact, so covers the actual variations that 
they become insensible to observation. 
Admitting that the heat in a column of air diminishes in the 
same proportion that the height increases; if we knew the rate of 
decrease, or the elevation necessary for depressing the thermo- 
meter one degree, we should be able to deduce the temperature 
at any altitude in the atmosphere from the temperature at the 
earth’s surface ;~and likewise to compute the height of a column 
of air, from having given the temperatures at its extremities. To 
find the rate, we must have recourse to experiments. But a very 
slight attempt to determine this quantity is sufficient to show that 
it is extreniely variable, even in circumstances in which it is im- 
possible to discern any apparent difference. In 38 measurements 
recorded by Ramond, made in circumstances very various, from 
observations free from the suspicion of great errors and leading 
Vol, 58, No, 279, July 1821, D to 
