PROFESSOE IvOPP ON THE SPECIFIC HEAT OF SOLID BODIES. 
91 
siderable for porous charcoal (in fact Pouillet’s experiments make this probable). 
This source of error is excluded in my method. 
20. In order to appreciate the trustworthiness of the results arrived at by my mode 
of experiment, it is important to state with what amount of accuracy the data of obser- 
vation and the ancillary magnitudes were determined. I will give this statement in what 
now follows. 
Por observing the temperature of the water in the calorimeter I used thermometers 
made by Geissler of Bonn, which the kindness of Professor Buff, Director of the Phy- 
sical Cabinet in Giessen, placed at my disposal. In these thermometers the tube 
consists of a fine glass thread drawn out at the lamp. The bulb is cylindrical, very 
thin in the glass, and contains but little mercury. On one ( b ) 1° C. corresponds to a 
length of almost 5 millims. on the scale, and on the other (r) to almost 4*5 millims. 
Tenths of a degree can be read off directly on the scale, and it is easy to learn to 
estimate hundredths safely. I have repeatedly compared these two thermometers, 
between 7° and 24°, with two normal thermometers of my own construction, which 
agree very well with each other, and on one of which a division corresponds to 0 o, 4878, 
and the other to 0 o, 4341. The differences of the indications between the Geissler’s 
thermometers and these could be considered as constant within those limits ; for the 
differences thus observed all the readings made with the Geissler’s thermometers had 
to be corrected to make them comparable with the indications of the normal ther- 
mometer. 
The temperature of the mercury-bath was ascertained by means of one of these 
normal thermometers, and the indications of this thermometer immersed in the bath 
(d in fig. 1.) corrected for the lower temperature of the mercury thread out of the 
bath ; this latter temperature was given with adequate approximation by the second 
thermometer, e. 
21. The weight of the thin sheet-brass calorimeter, together with stirrer, was 11T45 
grms.* Taking the specific heat of brass, according to Begnault, at 0-09391, the 
calorimetric value in water of this mass of metal is 1 -046 grm. Considering that the 
calorimeter in the experiments was not quite filled with water, but about ^th remained 
empty, even after introducing the tube, I put the value in water at 0-872. 
In determining the calorimetric water value of the immersed parts of the thermo- 
meters r and b, the following experiments were made. The weight of water in the 
calorimeter, together with the reduced weight of the metal, was 30-87 grms. When the 
thermometer r heated to 33 0, 86 was immersed, the temperature rose from 10°-73 to 
10 o, 85 ; the immersion of the thermometer b at a temperature of 37°-53 caused a rise from 
10°-61 to 10°-76. In both cases the temperature of the water was indicated by means 
* At the beginning of these investigations. During their progress the calorimeter was cleaned and dried 
with bibulous paper a countless number of times, so that its weight diminished by about 0-04 grm. in the 
course of the experiments. In determining the weight of water used in each experiment, the weight which the 
calorimeter actually had at the time was taken as basis. 
