134 



Mr Bevan, On the Joule-Thomson Effect. 



The following numbers are given by Joule and Thomson* for 

 the cooling effect divided by the difference of pressure in question 

 for air. 



•0170 for a difference of pressure 



•0179 



•0165 



0196 

 0177 

 0172 

 0174 



7 

 8 

 22 

 19 

 39 

 65 

 69 



The pressures are measured in pounds per square inch. 



The individual experiments on which these numbers are 

 founded differ very considerably among themselves, the variations 

 being in some cases very large. For example the cooling effects 

 from which the first number in the above table was deduced 

 varied from "094° C. to "125° C. For the fourth number in the 

 above table there were four observations, "358°, "348 , ^SO and 

 "373°. The best series of observations was the series for the last 

 number where the variations amounted to only about 6 per cent, 

 ■of the mean value. There is thus room for a considerable varia- 

 tion of the cooling effect in the difference of pressures in the case 

 of air. 



For carbon dioxide the numbers are in better agreement, being 

 for the cooling effect divided by the pressure difference : 

 "0770 for a pressure difference of 6 

 •0778 „ „ „ „ 18 



•0789 „ „ „ „ 37 



•0833 „ „ „ „ 60 



For the last number the temperature was about 70° C. lower 

 than for the others, and so the observation is not comparable with 

 those taken at nearly the same temperature. 



The observations for carbon dioxide are in much closer agree- 

 ment than in the case of air, but the number of comparable 

 observations being only three with a not very large range of 

 pressure difference, room is left for a considerable dependence of 

 the cooling effect on the pressure. 



With the experimental results at present available we cannot, 

 therefore, obtain any reliable approximation to the characteristic 

 equation for fluids. Until more accurate experiments have been 

 made by which the relation of the cooling effect to the pressure 

 of the gas can be determined, this method cannot lead to any 

 equation which will represent the results of direct experiment on 

 p, v, and 6. 



* Joule, Scientific Papers, Vol. n., p. 258. 



