538 M. V. Regnault on an Air-Thermometer 



grins, of mercury are added to this flask, which is then placed in 

 the furnace in the part the temperature of which is to be known. 

 The mercury soon boils, its vapour expels the air, which escapes 

 by the orifice o ; the excess of mercurial vapour emerges by the 

 same issue. When the apparatus has acquired the temperature 

 of the furnace, the plate c d is taken out in order to close the 

 orifice o, the flask is withdrawn and made to cool rapidly. The 

 mercury w r hich remains in the flask is weighed ; it is removed by 

 the addition of water and agitation. The mercury can be weighed 

 directly ; or if it contains impurities, it is dissolved in acid and 

 estimated as precipitate. 



The neck may also have the form of fig. 6. The tubulure 

 then terminates in a hollow conical part, and in its orifice is a 

 ball B of the same material as the flask. This does not close it 

 hermetically, but it prevents the currents of air of the furnace 

 from disturbing the mercurial vapour in the interior of the flask. 

 Let 

 V be the capacity in cubic centimetres of the flask at 3 as 



given by the weight of water which fills the flask ; 

 k be the coefficient of cubical expansion of the substance of 



which it is made ; 

 H be the height of the barometer at the moment at which it 



is withdrawn from the furnace ; 

 h the difference between the pressure in the interior of the 

 furnace and of the surrounding atmosphere. This differ- 

 ence may often be neglected, but it can readily be deter- 

 mined by means of a water manometer. 

 H the height H — h reduced to 0° ; 



S the theoretical density of mercurial vapour compared with 

 that of air under the same circumstances of temperature 

 and pressure ; that is, the density of this vapour taken at 

 temperatures starting from which it no longer changes ; 

 m other terms, higher than that, starting from which, mer- 

 curial vapour and air follow the same laws of expansibility 

 and compression. It is, moreover, clear that it is only above 

 this temperature that the pyrometer can give exact tempe- 

 ratures. 

 p the weight of the mercury which remains in the flask at the 



end of the experiment. 

 The weight of mercurial vapour which fills the flask at the 



maximum temperature x is 

 we have then 



V.*±^. 0-0012932. 8^ 

 1 + ux 760 



V.i±^ 00012932.8^ 

 1 -f ax i GO 



