354 Mr. J. Ro.se-lnne.s on the Attainment of 



accepted as the true explanation until it has been tested in 

 some way or other, since there is always the possibility 

 that a portion of the data not due to Joule and Kelvin is 

 likewise erroneous. 



The most hopeful method of attacking the question appears 

 to be to devise some way of arriving at an estimate o£ the 

 freezing-point using different gases, but without employing 

 the Joule-Thomson numerical data. One way of securing 

 this result is given in the following investigation. 



It was shown by Lord Kelvin that when a gas is forced 

 through a porous plug, we have 



\dt Jp ' 6p 



(Reprinted Papers, vol. iii. p. 179). 

 Divide by t'^ 



l/dv\ _ r _ JK^ 



t \dt)p t^~ t^ 8jr 



Integrate this equation with respect to t along an isopiestic 

 between the limits t^ and ti ; we thus obtain 



1 *o ~J,. t' h 



The symbols v^ and v^ denote the volumes at ^i and t^ for 

 the same pressure p ; if we integrate along a second isopiestic 

 2^ between the same limits of temperature t^ and ^i we shall 

 obtain 



It was shown by Joule and "Lord Kelvin that -^ is 

 independent of the pressure, heiice by subtraction ^ 



'^\-Vi vo — ro' ^Q 



which leads to 



Let the suffix 1 refer to the boiling-point, and the suffix 

 to the freezing-point ; then the above equation enables us to 

 calculate out the absolute value of the freezing-point. 



