600 REGNAULT ON THE 
_ (E—C) (C—B)—(A—C) (D—C) 
«P=(K—C) (C—B)—(A—B) (A—C)’ 
from which are deduced « and 8, and subsequently a, and 3. 
The values of the coefficients a, b, c are calculated afterwards 
by the formulz 
(a—B) (#1) ” 
_—Aa+(1+a2) B—C 
~  (@—B8) 8-1)” 
: a=A—b—c. 
A great number of determinations of the tension of aqueous 
vapour at 0° will be found in the preceding tables; but as 
perfect an accordance as I could have wished will not be met 
with in these determinations. It will be observed that the 
numbers obtained in the same series differ but very little; the 
errors of observation are therefore extremely small. The differ- 
ences only become slightly perceptible when one series of experi- 
ments is compared with another ; and in general when a differ- 
ence is perceptible at 0° it remains constant at all higher tem- 
peratures ; these differences therefore are attributable to con- 
stant errors, which I have in vain endeavoured to remove. I 
adopted in calculating the formula the number 4™°60 as the 
tension at 0°; it is the mean of the extreme observations 4°52 
and 4°68. 
The tensions corresponding to the temperatures 25°, 50° 
and 75°, were obtained by constructing with exactness, upon 
a large scale, the curve of the tensions, according to the results 
of the experiments, and correcting the numerical value of the 
ordinate taken upon the curve, by a little interpolation made 
upon the surrounding points, obtained directly from the obser- 
vations. This interpolation did not sensibly change the values 
of the corresponding ordinates for ¢ = 25° and ¢ = 50°. The 
graphic curve gave for ¢ = 75°, e = 288™™-8 ; the interpolation 
reduced this value to 288™™50, 
Setting out from the given numbers, 
° mm 
t= 0, e= 4-60 
t= 25, e= 23°55 
t= 50, e= 91:98 
t= 75, e = 288°50 
t = 100, e = 760°00 
