the Thermodynamic Sale of Temperature. Ill 



It is by no means essential to the success of this explana- 

 tion to suppose that the whole of the error o, 2 C. should be 

 attributed to M. Chappuis's experiments. Even if so small 





0°C' 



20 



30 



40 



50 



60 



7.) 



80 



90 



100 



a fraction as 0°05 C. be attributed to M. Chappuis, and the 

 remainder to the Joule-Thomson experiments, the above 

 method of reasoning would still apply. 



Theory of the Constant-Pressure Gas-Thermometer. 



TTe have discussed the theory of the constant-volume gas- 

 thermometer first, because this instrument has been largely 

 used by physicists in their thermal researches. The constant- 

 pressure thermometer is also employed sometimes, but it is 

 not generally deemed so trustworthy an instrument. Recent 

 improvements have gone iar to remove the practical objections 

 that were formerly attached to its use ; and we accordingly 

 propose to discuss its theory in this section. 



We start with the equation 



*.-*-,{*+§+ a}. 



Let us now imagine that we keep p constant ; let the suffix 

 applied to v and t refer to the freezing-point, and the suffix 1 

 to the boiling-point,, then 



^ =E '°-H ao+ 2r. + W' 



Multiply the last equation by i , and the last but one bv t x ; 

 we then obtain bv subtraction 



