202 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
than 650 cub. centims., and usually varied from 350 to 600 grammes. The weight of 
the flasks was taken empty, just previous to a determination of the flow and after 
they had stood for some time unstoppered inside the balance case. The interior of the 
flasks was always wet on account of the water 
which had been weighed in them on a previous 
occasion. 
The weight was taken on a large Oertling 
balance, which proved to be most suitable for the 
purpose. It was sensitive to less than 1 milli¬ 
gramme with 500 grammes, which gave a measure 
of the weight to a sufficient accuracy. The 
weights used were brass, and were one of several 
sets supplied us by Oertling. They were 
kept entirely for the present purpose. It is 
exceedingly unlikely that their errors would 
amount to as much as 1 milligramme, especially 
as the several sets sent us by Oertling agreed 
much closer than that, and the different weights 
in the same set agreed very closely amongst themselves. Even if it could be imagined 
that the sum of the errors of the weights used in any given weighing could have 
amounted to 10 milligrammes, that would have produced an error in the estimation of 
the flow of only 1 part in 50,000, whereas it is most probable that the error was not 
so much as a tenth of this. 
The correction to be applied to the weight for the ratio of the arms of the balance 
was found to be less than 1 in 100,000 for the weights used. 
In the reduction of the weight of the water in the flask to weight in vacuum, it is 
necessary to correct for the presence of water-vapour in the displaced air. This water- 
vapour is retained in the calcium-chloride tube when the air is driven out and 
therefore appears not only in the weight of the flask empty, but when it contains the 
weight of water. It is consequently eliminated from the final weight. The actual 
weight of air displaced, however, is less than it otherwise w r ould be, by the presence 
of the water-vapour. In applying this correction it may be assumed that the air inside 
the flask is completely saturated with water-vapour at the temperature of weighing. 
Where brass weights are used the ratio of the weight in vacuum to the weight in 
air is given by the expression 
W vac. _ / \t_ \ t \ 
W air ~ \ + S “ 8 4/’ 
where for the calculation of X t we can use the formula 
\ __\_ P P 
1 ~ (1 +’-003660 T ‘ 
