ATTAINMENT OF VERY LOW TEMPERATURES. 9 



atmospheres flowed in a steady stream through a phig- of compressed 

 fiber so that the increase in the kinetic energy of the gas, which would 

 take place if the gas streamed through an orifice in a plate, was re- 

 duced to a minimum, and practically eliminated. In passing from the 

 higher to the lower pressure work was done in overcoming internal 

 friction, so that were the gas a perfect one its total energy content 

 would remain constant, and there would be no change of temperature. 

 In the case of air and carbon dioxide it was found, however, that a fall 

 of temperature took place ; hydrogen, on the other hand, became 

 warmer. 



The change of temperature must be attributed to the performance 

 of work in overcoming internal stresses ; though as to what the nature 

 of these stresses may be we have at present no knowledge. In 

 many text-books of physics it is stated that the direction of the change 

 of temperature, the sign of the Joule-Thomson effect, is connected with 

 the variation of the product '' PV " with pressure for the gas. Though 

 for hydrogen (and helium) the value of the product increases with 

 rise of pressure, while it decreases in the case of air and all other gases, 

 this relationship does not really apply. For pressures above fifty 

 atmospheres the value of the product PV for air increases with rise of 

 pressure, yet, if we expand air from loo atmospheres to 50 atmos- 

 pheres, it still becomes cooled. 



For small differences of pressure the change of temperature was 

 found to be directly proportional to the difference between the pressure 

 on the two sides of the plug. The results are stated by the original 

 investigators in terms of degrees Centigrade per hundred inches of 

 mercury differences of pressure ; it is now more usual to express the 

 value of the Joule-Thomson effect in terms of degrees per atmosphere 

 difference of pressure. 



The experiments which were carried out at temperatures between 

 the boiling and melting points of water, showed that the magnitude of 

 the effect varied with the temperature, and might be represented by the 

 formula, 



where / is the Joule-Thomson Effect, T is the absolute temperature 

 and a and /5 are constants. The values of the constants for air and 

 hydrogen are as follows : 



a P 



Air 64.1 —0.331 



Hydrogen 441. 5 0.697 



These figures indicate that for either gas above a certain tempera- 

 ture, the inversion temperature of the Joule-Thomson effect, the effect 



