592 DE. A. DUPRE AND F. J. M. PAGE ON THE SPECIFIC HEAT AND OTHER 
stretched silk cords and surrounded by a double cylinder of tin plate, to prevent as far 
as possible any gain or loss by radiation. 
A thermometer, divided into tenths of a degree Centigrade, allowing yjjo of a degree 
to be read, was used to indicate the temperature of the steam-oven. 
Two thermometers were employed to indicate the temperature of the calorimeter, — one 
having a range from 10° to 20° C., and divided into twentieths of a degree Centigrade, 
allowing of a degree to be read off by means of a telescope ; the other had a range 
from 16° to 21°*5 C., divided into fiftieths of a degree C. and allowing 5 ^ of a degree 
to be read off. 
The bulbs of both these thermometers were thin, and of such a length as to pass through 
nearly the entire depth of liquid contained in the calorimeter. A thermometer similar 
to the two last was employed to take the temperature of the air. All the thermometers 
were carefully compared with a standard, as well as calibrated by thread measurements. 
The experiments are conducted as follows. As soon as the temperature of the steam- 
oven remains constant for at least ten or fifteen minutes, it is read off ; a measured 
quantity of the liquid, cooled a few degrees below the temperature of the room, is 
poured into the calorimeter, and its temperature, after thorough mixture, read off by 
means of a telescope. On a given signal, one assistant pushes the calorimeter under- 
neath the steam-oven ; a second then, opening the slide-valve, lets down the weight which 
is detached from the string, and the calorimeter is brought back into its first position. 
The weight, which has of course been kept underneath the surface of the liquid after 
leaving the steam-oven, and has not been allowed to touch the sides or bottom of the 
brass vessel, is now hooked to a strand of worsted, previously twisted ; on allowing this to 
untwist a rapid rotation of the weight ensues, and consequently a thorough mixture of 
the liquid by the fan-wheel*. The whole of these movements occupy from fifteen to 
twenty-five seconds. 
The elevation of the temperature of the calorimeter is observed, which usually reaches 
a maximum one minute after the introduction of the brass weight, whilst with the 
copper weight a minute and a half elapsed before this was effected. 
As soon as the temperature has attained its greatest elevation and the thermometer 
begins to sink, the time is noted ; and in half a minute with the brass weight, or three 
quarters of a minute with the copper weight, the temperature is again read off ; the fall 
of temperature during this interval is added, as a correction, to the highest temperature 
observed, as representing the loss due to radiation. This correction usually amounts 
only to a few hundredths of a degree. 
* As a mean of several experiments, it was found that a rapid rotation of the weight during five minutes 
produced a rise in the temperature of the water in the calorimeter of 0°-005 C. In the course of an experi- 
ment on the specific heat of a mixture, the time during which the weight rotated at the above speed always 
fell short of one minute, consequently the effect thereby produced on the temperature of the calorimeter is 
somewhat less than 0 o, 001 C. The heat thus produced is moreover a nearly constant quantity, affecting all 
experiments alike, so that its ultimate influence on the specific heat of the various mixtures is reduced to an 
almost infinitesimal fraction; it has therefore been neglected throughout the calculations. 
