602 REGNAULT ON THE 
melting; the reservoirs were blown upon the stems by the same 
workmen, and nevertheless these thermometers did not present 
so complete an accordance in their indications as might rea- 
sonably have been expected, from the extreme care with which 
the divisions had been made and their scales tested. This ob- 
stacle is overcome by observing the temperatures above 100° 
directly by an air thermometer ; but to render this latter instru- 
ment susceptible of a certain amount of exactitude, it ought to 
be of considerable size, in order that the tubes, in which the 
changes of the elastic force of air are measured, may have so 
large a bore, as to affect but insensibly the height of the liquid 
contained in them by capillary action. But such an instru- 
ment could not be placed in my apparatus. I preferred there- 
fore to observe in a direct manner the temperatures on mer- 
curial thermometers, and to compare afterwards these same 
thermometers with an air thermometer constructed in the proper 
manner, and placed in similar circumstances to those, in which 
these instruments are placed in my retort. J have already made 
several attempts to make this comparison, but the results are 
not satisfactory ; I propose therefore to make fresh experiments 
and to publish them with the remainder of my observations on 
the elastic force of aqueous vapour at high temperatures. 
In the mean time I thought it better to connect the observa- 
tions made up to this time for high temperatures by a provisional 
formula of interpolation, which enables us easily to study their 
course. I adopted simply the formula with two terms, 
loge = a+ ba’, 
The three constant quantities in this formula were calculated by 
the following values, 
° mm 
t= 100 e= 7600; 
t= 193 e = 16210; 
t= 146 e= 3177-0; 
which were taken on the graphic curve constructed from the 
observations of the series R. 
Thus is obtained 
loga= 1:9977641 
log = 0:4692291 } . 
a = + 5°8267890 
The formula is 
loge = a — ba’, v=t—100°.. . . . « (GC) 
