332 Dulong and Petit on the Measure of Temperatures, [May, 
nature of the gas ; for our thermometer m is equal to 0-00919 
in air; to 0°0318 m hydrogen; to 0:00887 in carbonic acid; 
and to 0:01227 in olefiant gas. (These values of m suppose p 
expressed in metres, and ¢ in centigrade degrees.) We may, by 
the preceding value of V, calculate the ratios of the cooling 
powers of the different gases for each pressure. Thus taking the 
cooling power of air at unity, and supposing the pressure 
= 0°76 m. we have for the cooling power of hydrogen 3-45, and 
for that of carbonic acid 0-965. These numbers will change 
with the elasticity belonging to the three gases. This Messrs. 
Leslie and Dalton did not perceive ; but it is easily deduced 
from our formula. However, their determinations differ but little 
from those which we have calculated for the pressure of 0°76 m. 
We should deduce likewise ratios very little different from these 
from the experiments made more recently by Sir Humphry 
Davy. 
The simplicity of the general law which we have just made 
known made us desire eagerly to be able to verify it at tempera- 
tures more elevated than those which we had attempted in our 
experiments. We succeeded by a very simple process, the idea 
of which was first suggested by Mr. Leslie. 
When our thermometer with the naked ball cooled in the 
open air, the total velocity of this cooling is the sum of the velo- 
cities due separately to the contact of air and to radiation. 
Denoting these by v and v’, the total velocity is v + v’. If the 
thermometer be covered with silver, the velocity ¥ due to the 
air remains the same for the same temperature, and v’ is reduced 
to Moe since the constant ratio of the radiating powers of 
glass and silver is 5°707. The total cooling of the silvered 
thermometer is then v + sr Hence it is easy to conclude 
that in order to know at all temperatures the losses of heat pro- 
duced by the contact of air, it is sufficient to determine the 
total velocities of cooling of our thermometer, first when the 
bulb is naked, and then when it is covered with silver. These 
velocities being represented by a and 0, we shall have 
a=v+u' b=v+ sao 
Po a A 5101 xb —a@ 
ra A107 
Let us apply this formula to the results contained in the fol- 
lowing table : 
Vag 
