[ 722 ] 



LXIV. Temperature of Inversion of the Joule -Kelvin Effect 

 for Air and Nitrogen. Preliminary Communication. By 

 K. Olszewski*. 



IN a paper published by me five years agof, I bad de- 

 termined the inversion temperature of the Joule-Kelvin 

 effect to be —80°' 5 for hydrogen; this figure has proved to 

 be of fundamental importance in the design of liquefying 

 apparatus^ for this gas. The paper mentioned has also 

 attracted the attention of physicists, and A. W. Porter § has 

 made use of it in a theoretical investigation of the accuracy 

 of van der Waals's and Dieterici/s state equations, based on 

 my determination of the inversion temperature. On account 

 of the great theoretical importance of such determinations, I 

 have decided to carry out similar measurements for other 

 gases, and above all for air and its more important com- 

 ponents. Up to the present, I have only completed the experi- 

 ments on the inversion temperatures for air and atmospheric 

 nitrogen, and take the opportunity of presenting the results in a 

 brief notice, reserving a detailed description of the experi- 

 mental arrangements and the apparatus for a later communica- 

 tion. I only wish to remark that the apparatus used did not 

 differ in principle from that employed by me five years ago, 

 but that in view of the high temperature (up to 300°) at 

 which the experiments on air and nitrogen had to be carried 

 out, considerable modifications in the details as well as in the 

 dimensions became necessary. 



As Witkowski, as far back as 1898 ||, and, last year, 

 Porter §, on theoretical grounds came to the conclusion that 

 the inversion temperature of the Joule-Kelvin effect for gases 

 was a function of the pressure, I have in the present experi- 

 ments devoted special attention to the initial pressures of the 

 gases which were made to undergo an irreversible expansion. 



The apparatus was heated in an oil-bath. Temperatures 

 were measured with a high-reading mercury thermometer ; 

 but for the purpose of determining the small temperature 

 changes which occur during expansion, an iron-constantan 

 thermo-couple was used, whose sensitiveness was about 0°*2 

 per mm. of galvanometer scale. 



The gas, which had been compressed to the initial pressure 

 p, was allowed to expand down to one atmosphere. The 

 experiment was repeated a number of times under these con- 

 ditions, the temperature of the gas being gradually reduced 



* Communicated by the Author. (Bull. Acad. Crac. Nov. 1906.) 

 t K. Olszewski, Bull. Ac. Crac. 1901, p. 453 ; Phil. Mag. [3] p. 535 (1902). 

 % K. Olszewski, Bull. Acad. Crac. 1902, p. 625, and 1903, p. 241. 

 § A. W. Porter, Phil. Mag. [6] xi. p. 554 (1906). 

 || A. W. Witkowski, Bull. Acad. Crac. 1898, p. 282. 



