228 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
referred to conduction from the outflow tube, which was the same size in both 
calorimeters. 
The observations under February 22 and March 2 were taken with Calorimeter C, 
using the limits of flow which I have since adopted for the present measurements. 
Calorimeter C.—Mean Temperature, 29°-ll C. February 22, 1899. 
8 = -00490. 
J = 4-1794. 
eld. 
Q- 
(EC - JQ d0)/dd. 
Difference from 
mean, '04938. 
8-2680 
•392606 
•04939 
+-ooooi 
8-2635 
■392663 
•04937 
- - ooooi 
8-1938 
•496708 
•04928 
- -oooio 
8-1844 
•496591 
•04946 
+ -00008 
7-9031 
•660865 
•04945 
+-00007 
7-9083 
•658741 
•04932 
- -00006 
Calorimeter 
C.—Mean Temperature, 29 0, 21 C. 
March 2, 1899. 
8 = — -00499. 
J = 4-1790. 
dd. 
Q. 
(EC - JQ dd)/dd. 
Difference from 
mean, -04968. 
8-4310 
•375154 
•04963 
- -00005 
8-4304 
•375076 
•04967 
- -ooooi 
8-3979 
•472489 
•04967 
- -ooooi 
8-4060 
•471670 
•04986 
+-00018 
8-3390 
•590477 
•04960 
- -00008 
8-3439 
•589356 
•04969 
+ -ooooi 
The agreement of the heat-loss is very satisfactory, and the variations from the 
mean value are easily within the limits of error of all the measurements, and are all 
less than 1 part in 10,000. 
The three sets of readings for Calorimeter C show a small difference between the 
values of the mean heat-loss. This shows, as has been pointed out, that the absolute 
value of the radiation loss for one calorimeter cannot be relied on from time to time, 
but will vary, for many reasons. Ffowever, this never produced any error in the 
measurements of the specific heat of the water, on account of the method adopted of 
always eliminating the heat-loss from at least two different flows taken within a 
