Notices respecting New Books. 55 



by any body in given circumstances. This absorption of heat is 

 only to be perceived by the changes which take place in the state 

 of the body or by its dilatation. Now when the dilatations which 

 different bodies undergo under identical circumstances are studied 

 comparatively, it is soon seen that these dilatations are far from fol- 

 lowing the same law. As to the quantities of heat acquired by the 

 different bodies, when they are brought successively to different 

 temperatures measured by the dilatations of any one of them, it is 

 perceived that they are variable and unequally variable in each, 

 while no one has yet succeeded in determining the relations which 

 exist between these variations of capacity and these variations of 

 volume. 



It is evident therefore how difficult it would be to construct a 

 perfect thermometer, that is to say a thermometer whose indications 

 would always be proportional to the quantities of heat it absorbed, 

 or in which the additions of equal quantities of heat would always 

 produce equal dilatations. However, physicists imagined that they 

 had found this normal thermometer in the gas-thermometer. This 

 opinion was founded on the following considerations, namely, that if 

 thermometers are made with air and several other solid or liquid 

 substances, and graduated between 0° and 100°, with the supposi- 

 tion that the dilatation is uniform, when this graduation is prolonged 

 beyond 100°, the air-thermometer is that which, for temperatures 

 above 100°, will constantly indicate the lowest temperature; and 

 while the oth-ers will indicate all the temperatures different from and 

 higher than that which the gas-thermometer denotes, the latter will 

 denote the same, whatever be the nature of the gas, whether, for 

 example, it be air or hydrogen. This result of observation seemed 

 moreover to be confirmed, by the opinion that was entertained that 

 gases are subject to simple and general laws, and by the notions 

 that were deduced from it as to their physical constitution, in which 

 no part was attributed to the proper nature of the atoms, and which 

 was made to depend solely upon heat, an agent common to all. 



But the researches of M. Regnault, by demonstrating that the 

 laws which regulate the constitution of gases are far from being so 

 simple and general as had been believed, have rendered it inadmis- 

 sible to regard the gas-thermometer as a normal thermometer. The 

 indications of the gas-thermometers. Hive those of other thermometers, 

 can therefore only be considered as more or less complicated func- 

 tions of the quantities of heat. 



In default of giving an exact measurement of the quantity of heat, 

 the thermometer must at least always remain accurately comparable 

 with itself, that is, must always furnish the same indication in iden- 

 tical conditions : and, further, it is necessary to be able to reproduce 

 at will and always obtain instruments rigorously comparable. Now 

 the mercurial thermometer which fulfills the first condition well, only 

 fulfills the second very imperfectly. Thus two mercurial thermo- 

 meters, regulated for the same fixed points of melting ice and of the 

 boiling of water under the pressure of 0*7t>0 millira., may present very 

 considerable differences in the course beyond these fixed points, if 



