612 DR. t. s. humpidge on the atomic weight of glucinum. 
and if possible to modify the apparatus so that it can be used for higher temperatures 
than 100°. 
The specific heat of the sample of turpentine was next determined, with results as 
follows, T, t, 9, and t' having the same significance as before, W being the weight of 
turpentine used, w the weight of silver, s its specific heat (taken as 0"057), and s' the 
required specific heat of the turpentine. 
Silver in turpentine. 
I. W=75-30, io= 10-205, T=102°-2, <=10°-98, 0=12°-47, t'= 12 0, 8. 
m»( T —0)—2-91(fl -Q_ 
W (6-t) 
s = 
= 0-4265. 
II. W=75T9, io= 10-205, T=102°-5, t= 13°*32, 0=14°*81, T=14°-5. 
s'=0’4165. 
III. W= 74-74, ic= 10*205, T=102 Ol 4, *=14°-56, 0=16°-OO, «'=15°*0. 
s' = 0-4280. 
Mean specific heat = 0'4236; mean error = 0'0044 = 1 per cent. 
The result obtained by Regnault for turpentine at 10° was 0"4278, and by Hirn 
0-4241. 
Three determinations of the specific heat of the compressed disc of glucinum referred 
to above were then made, yielding the following closely concordant results, in which 
the letters have the same significance as before, except that W is the weight of the 
turpentine and W' its water equivalent + the water equivalent of the calorimeter, &c. 
Glucinum in turpentine. 
I. W=72-93, W'=33-77, w=0-6575, T=102°T, «=ll°-09, 6= ll°-85, T=ll°-1. 
W '(0-0 
s= 
w(T—9) 
= 0-4326. 
II. W=72-67, W=B3’66, iv= 0'6568, T=102°-2, £=10°*08, 0=lO°-86, t!= 9 0> 9. 
s=0-4264. 
III. W = 75*12, W' = 34-69, w=0-6565, T=102°-3, <=14°‘26, 9= I4°‘98, i'=14°'3. 
s=0-4357. 
Mean specific heat of glucinum =0*4316; mean error = 0*0035 = 0*9 per cent. 
