ELASTIC FORCES OF AQUEOUS VAPOUR. 601 
we obtain 
log’e,= 0006865036, 
logB,= 1°9967249, 
log b= 2°1340339, . 
loge = 0°6116485, 
a = + 47384380. 
The tables show the accordance which exists between the re- 
sults deduced from this formula, and those obtained by direct 
observation. I endeavoured to apply the same formula to tem- 
peratures below 0°, but I found that the elastic forces deduced. 
from it were constantly higher than those established by obser- 
vation. The difference is very small for the first negative de- 
grees of the scale, but increases up to 0™™25 towards—25° ; 
so that the curve deduced from the formula rises on leaving zero, 
and separates gradually from the curve observed for temperatures 
below 0°. 
The same formula cannot therefore be employed for tempera- 
tures much below zero; I have calculated for these a little for- 
mula of interpolation, 
(A.) 
e=a+ ba’, EPPS PIORRERS LL 
in which # = ¢ — 32°. 
The three constant quantities were determined by the three 
following values : 
mm 
= — 32, a= 0; é = 0°31 5 
‘t= — 16; w=" 16; ee 1°16 5 
t= 0, @ = 32; e = 4°60. 
In this manner was obtained 
log b= 1:4724984, 
log # = vosrises | decir ias et 2(13,) 
a@ =+0°0131765. 
The temperatures were measured in all my experiments by a 
mercurial thermometer, which can be much more promptly and 
precisely read off than an air thermometer; but beyond 100° 
mercurial thermometers become inexact; they vary for the 
same temperature according to the nature of the glass of which 
the reservoir is constructed, and probably according to the 
manner in which the reservoir has been blown. The four ther- 
mometers which were employed in my experiments below 100°, 
were constructed with tubes of crystal all coming from the same 
