172 ANNUAL EECORD OF SCIENCE AND INDUSTRY. 



best length for this spring. According to bis experience, 

 the length of the spring, and the length alone, is sufficient 

 to secure perfect isochronism. Ilorolog. Journal, June, 1875. 



THE VARIATIONS OF TEMPERATURE ACCOMPANYING THE DIF- 

 FUSION OF GASES. 



Professor Dufour, of Lausanne, Switzerland, as the results 

 of an investigation into the variations of temperature which 

 accompany the diffusion of gases traversing partitions of 

 porous earthenware, states his conclusions as follows : AVhen 

 currents of dry air, of hydrogen or of illuminating gas, 

 circulate along the walls of a porous vase, or of a vase 

 which incloses fragments of porous material, they produce a 

 lowering of temperature. The depression diminishes little 

 by little, and finally ceases altogether. When the cur- 

 rents of the same gas, charged with moisture, circulate 

 under the same conditions, there is produced a heating, 

 which also diminishes gradually, and finally ceases. The 

 warming and the heating are more or less considerable, ac- 

 cording to the initial condition of the porous vase. The 

 greatest variations are produced when the dry current suc- 

 ceeds to a saturated current, or inversely. These variations 

 of temperature are probably due to the absorption of aque- 

 ous vapor by the porous substance, or to the disengage- 

 ment of this vapor. If the experiments are conducted 

 under a constant barometric pressure, then, when the air on 

 the one side, and the hydrogen or illuminating gas on the 

 other side, are in contact with the two faces of the porous 

 partition, the diffusion which takes place produces a chaYige 

 of temperature, but a change having a different sign on the 

 opposite sides of the diffusing partition. There is a lowering 

 of temperature on the side where the denser gas is found, 

 or, in other words, on the side where the current arrives 

 most abundantly. There is, on the other hand, a rise of tem- 

 perature on the opposite side. These variations of tempera- 

 ture have been observed when the gases taking part in the 

 diffusion are dry, as well as when they are charged with 

 aqueous vapor. When the gases are employed without 

 drying, and without saturation, the diffusion also evidently 

 occasions the variations of temperature just indicated ; but 

 it is probable that this variation is influenced by the pres- 



