56 Notices respecting New Books. 



the envelopes of these thermometers are not formed of glass of the 

 same nature, or the molecular condition of which is not the same. 

 In fact, in any thermometer formed by a liquid or gaseous substance, 

 the indications of the instrument depend upon the dilatation of that 

 substance, and of that of the envelope. The dilatation of mercury 

 being only seven times greater than that of the glass which encloses 

 it, the variations presented by the different dilatations of different 

 kinds of glass form very sensible fractions of the apparent dilatations 

 of the mercury, and consequently influence notably the indications 

 of the instrument. In the gas-thermometer, on the contrary, the 

 dilatation of the gas being a hundred and sixty times greater than 

 that of glass, the variations in the dilatations of the different kinds 

 of glass do not sensibly influence the indications of the apparatus, 

 and do not prevent the comparison of the instruments. At the 

 same time, it is important to determine the conditions in which this 

 instrument remains comparable. 



Thus in the question of the measurement of temperatures, it is no 

 longer sought to obtain an instrument which measures the quantities 

 of heat, or the indications of which will be proportional to the tem- 

 peratures. The pretension of the physicist is more modest ; he asks 

 only the possibility of constructing instruments which, in the same 

 conditions of temperature, shall always give exactly the same indi- 

 cation, that is, which shall be comparable with each other and with 

 themselves. The study of this desideratum is the object of that one 

 of M. Regnault's memoirs which treats of the measurement of 

 temperatures. He first discusses the gas thermometer, then the 

 mercurial. Finally, he devotes an important part of his memoir to 

 the measurement of temperatures by means of thermo-electric cur- 

 rents. We shall give in a future number, almost word for word, this 

 last portion, which is very little known, and in which M. Regnault 

 treats several interesting questions in thermo-electricity, with that 

 same rigorous precision which he brings into all his researches. At 

 present we shall confine ourselves to a rapid survey of the first two 

 portions of the memoir, devoted to the gas and the mercurial ther- 

 mometers, which form a whole, the more complete, that in regard to 

 the measurement of temperatures, M. Regnaul thas been led to re- 

 ject the employment of thermo-electric currents entirely, and that, 

 consequently, the interest of that portion of his researches lies wholly 

 in the details, and beside the principal subject. 



Gas TJiermometers. 



There are two ways of employing gas as a thermometric substance. 

 It may be placed in such conditions that the pressure to which it is 

 subject remains constant, its increase of volume being observed, or 

 the gas may be compelled to remain in the same volutne while we 

 observe the increase of its elastic force. 



The first method requires the employment of a capillary tube, 

 uniting a calibrated tube to the reservoir filled with air, which is 

 exposed to the temperature to be measured. This arrangement 

 allows of the calibrated tube being removed to a distance from the 



