180 Dulong and Petit on the Measure of Temperatures, [Maren, 
this effect, if it were to take place, would increase almost as the 
loss of heat of the body, so that the error produced would affect 
equally all the results. It was easy to raise the temperature of 
the surrounding medium, by passing vapour into the water 
through the tube S’ U V, plunging to the bottom of the liquid. 
The orifice of the balloon is shut by a thick plate of glass, 
A B, ground with the greatest care upon the edge of the balloon 
itself. The surfaces in contact have besides, in consequence of 
the thickness of the neck, a sufficient extent, so that the interpo- 
sition of a small quantity of hog’s lard renders the contact very 
close, and prevents all communication from without. 
This plate is perforated at its centre by a circular opening, into 
which a cork is firmly put, which contains the tube of the ther- 
‘mometer ; and the intermediate tube, CO, is of such a length 
‘that the bulb is preciselyin the centre of the balloon. By giving 
this intermediate tube a very small diameter, the quantity of 
mercury without the bulb is diminished, and the swelling which 
takes place at the commencement of the scale enables us to fix 
the tube more firmly in the cork. Thus the thermometer is fixed 
in the plate, and this disposition is shown particularly in fig. 6, 
where the bulb of the instrument is placed above the furnace, 
which serves to heat it. The screens, A A’, are leaves of tin 
plate, separated from each other, which serve to screen the 
plate, A b, from the action of the heat. 
Let us now return to fig. 5. The stem of the thermometer, 
which is without the balloon, as is evident from the figure, is 
covered by a hollow tube, 8 T, the ground bottom of which is 
applied to the upper surface of the glass plate. This kind of 
vessel is terminated above by a stop-cock, to which is cemented 
the end of the very flexible leaden tube, DEF. The other 
extremity of this tube is firmly fixed to the plate of an air-pump, 
HK. The canal, which in this machine makes the communi- 
cation between the centré of the plate and the barometer, is 
connected with another tube with a stop-cock, to which is 
cemented a tube filled with muriate of lime. It is through this 
tube that the gas passes by way of the bent tube, m npr s. 
The glass air holder, being moveable up and down, enables us to 
make the elasticity of the gas introduced the same as that of the 
atmosphere. We shall now describe our mode of proceeding in 
each experiment. 
The water in the trough being brought to the requisite tem- 
perature, and the thermometer fixed in the glass plate being 
heated to nearly the boiling point of mercury, it was transported 
rapidly into the balloon. ‘the glass, S T, already cemented to 
the leaden tube, was then drawn down over the stem. While 
the surfaces in contact were carefully luted, an assistant rapidly 
exhausted the balloon, by means of the air-pump. The commu- 
