136 
DE. J. P. JOULE ON THE SUEEACE-CONDENSATION OF STEAM. 
The following is my method of experimenting. The nut n being unscrewed, the 
dish of mercury removed from under the gauge-tube, and the water being completely 
discharged from the tap W, the cock T is partly opened, and the steam is blown through 
the steam-pipe s, the gauge Z», and the receiver E- until they are completely fi’eed from 
air. The nut n is then screwed on, W closed, and the water let on, the three operations 
being performed as simultaneously as possible. At the moment when the steam is 
about to cease issuing from the gauge-pipe, its end is introduced into the dish of mercuiy. 
After an interval of time, varjdng from half a minute to three minutes, the condensation 
goes on with perfect regularity, and the mercury in the vacuum-gauge remains steady. 
The temperature of the water flowing away and the gauge are observed every two or 
three minutes. The experiment is terminated by simultaneously shutting ofi” the steam 
and the water, and opening the tap W to let off the water remaining in the pipe. The 
nut n is then removed, and a quantity of air having entered the receiver, the condensed 
water is caught by a small can (held close and containing a thermometer), which over- 
flows into a larger vessel in which the water is immediately afterwards weighed. 
The values of several small corrections which had to be applied to the observations 
were obtained from data derived from separate experiments. Of the thermometers 
employed, one was made by Fastee, in which each division is equal to 0°‘225 ; the two 
others were from Kew Observatory, and have for each division the values 0°T and 0°‘0994 
respectively. A correction had generally to be applied in consequence of the non-immer- 
sion of the stems. 
The cooling effect of the atmosphere on the receiver E operates partly to condense 
steam and partly to cool condensed water. The correction on the former account was 
found to be equal to the continual product of the time in minutes, the proportion of 
acting surface, and the difference between the temperatures of the receiver and atmo- 
sphere, divided by 77 times the difference between 640 and the temperature of the con- 
densed water : the result had to be subtracted from the weight of condensed water. The 
correction on the latter account is equal to the continual product of the time, acting 
surface, and difference of temperature, divided by 77 times the weight of condensed 
water ; it had to be added to the observed temperature of the condensed water. 
The correction on account of the cooling of the refrigerating water on flowing through 
C into the vessel U, was found to be equal to the difference of temperature between the 
water and the atmosphere, multiplied by 0'61, and divided by the quantity of water 
flowing per hour. This rule applies to the case in which the external pipe C was 4 feet 
long and 1 inch in diameter. Corrections in the instances in which other tubes were 
used were made by calculation without express experiments, inasmuch as they were of 
very trifling amount. 
The slight loss of water by evaporation, before and during the process of weighing, 
was allowed for in the w^eighing both of the refrigerating and condensed water. 
The metal of the steam-pipe and receiver is necessarily at 100° at the commencement 
of an experiment, and therefore communicates some heat during the first few moments. 
