1170 PROFESSOR M. BELL ATI AHD DR. R. ROMANESE OH THE SPECIFIC 
brass cylinder was introduced into another very long brass cylinder, heated externally 
and at the lower end by water or paraffine, and closed by a cork. The stem of the 
thermometer, passing through this cork, supported at a convenient height the inner 
cylinder. 
This manner of heating the substance is, indeed, somewhat slow; but we adopted 
it because it is very easy to maintain the temperature constant for a long time, and 
because it allowed us to carry the substance, protected by the cylinder, from the 
heating apparatus to the calorimeter, in which it was allowed to fall, without any 
appreciable loss of heat. 
We used a Geissler’s thermometer graduated to C. ; but the temperatures 
were referred to the air-thermometer after having accurately compared our Geissler’s 
thermometer with a Jolly’s air-thermometer. 
The calorimeter (fig. 2) was a copper cylinder containing about 125 grms. of water, 
furnished with a movable co’ver, and was contained in a tin cylinder surrounded by 
water. For stirring the water a horizontal frame of copper wire covered by a brass 
net was moved up and down by a handle. The substance fell upon this net, and was 
moved in the water with the stirrer. 
The thermometer in the calorimeter was graduated to jf$° C., and was accurately 
compared with a standard thermometer. The thermal capacity of the calorimeter, 
stirrer, and thermometer was 2*660 grms. Calorimetric experiments and calculations 
were made according to Begnatjlt’s method, improved by Wullner.* In order to 
diminish corrections the temperature of the water in the calorimeter was initially 
cooled below the outer temperature. 
We have not described the physical properties of the bodies we have studied, because 
Mr. Bod well has done so in his papers. We have only given the temperature range 
of the contraction and expansion by heat of each substance, derived from Mr. Bod- 
well’s private communications, and afterwards the results of our calorimetrical 
experiments. 
Iodide of silver, AgL 
From 0° to 142° 0., slight contraction. 
„ 142° „ 156°'5 0., great contraction. 
„ 156 0, 5 „ 527° C., expansion. 
The calorimetric results are given in the following table, in which Q denotes the 
number of calories given out by unit weight of the substance in cooling from T to f C. 
The rate of change of the specific heat of water for temperature is no doubt small, but 
still uncertain, and therefore we did not introduce any correction for it; but we give 
in the table the initial water temperature r, so that the eventual correction is always 
possible. 
* A. WurLNEK, Wiedemann, Ann. (1880), vol. x., p. 284. 
