26 Remarks on the 
to results sufficiently exact, the quickest decrease of heat is at 
the rate of 61 fathoms, and the slowest, at the rate of 136 fa- 
thoms, to a centesimal degree. The mean rate deduced from all 
the 38 observations is almost exactly 90 fathoms; and we see 
that the extremes are different from the mean quantity, not by 
a small part of it, but by a half. Now, as we cannot doubt that 
the principle which distributes the difference of temperatures 
equally through the whole height of the column is nearly true, 
we must infer that the rate of decrease depends in a great mea- 
sure upon circumstauces peculiar to each particular case. In 
this respect causes seem to operate, which the observer is not 
only unable to appreciate, but even of the existence of which he 
has no indications. Very little confidence can therefore be placed 
in temperatures at different heights in the atmosphere, estimated 
by the rate of the decrease of heat; although the exactness of 
barometrical measurements is not by this means affected, the 
heat of the column of air being always determined by the tem- 
peratures actually observed at its extremities. 
Perhaps we may find, in the nature of the instrument with 
which the heat is measured, some reason for the irregular devia- 
tions of the observed temperatures of the atmosphere from any 
theoretical law. ‘The thermometer measures the temperatures 
only of such bodies as are in immediate contact with it. Local 
circumstances, impossible to be appreciated, may therefore so 
much affect a thermometer placed at the extremity of a column 
of air, as to make it indicate a temperature very different from 
what would take place at a medium, and when all the causes that 
influence the propagation of heat through its whole length, have 
produced their due effect. In this manner the observed may di- 
verge from the true temperature by a current of air in which the 
thermometer is placed; by the reflection of the sun’s rays from 
the neighbouring objects; by evaporation and the radiation of 
heat depending upon the nature of the soil in the vicinity, and by 
other causes. 
The rate of the decrease of heat deduced from the above-men- 
tioned observations of Ramond; that is, 90 fathoms to a cen- 
tesimal degree, or 100 yards to one degree of Fahrenheit ; seems 
to be the quantity most generally adopted. By means of this 
proportion, the temperatures that prevail at given altitudes in the 
atmosphere are sometimes determined with great precision, al- 
though in other cases the calculation is wide of the truth. Thus, 
taking the great height of 8817 fathoms ascended by Gay-Lussac 
in a balloon, the difference of temperature, at the rate of 90 fa- 
thoms to a degree, will be found 42°4, which is a near approxi- 
mation to 40°-3, the observed quantity. On the other hand, if 
we apply the same rule to the extreme cases in the Table of Ra- 
. mond, 
