584 



NATURE 



[April 1%, 1889 



atmospheres) the relation between these two quantities 

 may be expressed in all three cases by the general equa- 

 tion ^7/ = a -{• bp-\-cp'^. The curves which this equation 

 represents have their concave side towards the axis of p. 

 Between the temperatures - io3°'5 and 100", then, nothing 

 is observed that contradicts former experience with regard 

 to hydrogen on compression. 



But at the temperature of boiling oxygen, - 183', the 

 behaviour of hydrogen is very different. Up to a pressure 

 of about fourteen atmospheres the product/?/ decreases 

 with rising pressure, and the above equation no longer 

 holds good. At about fourteen atmospheres, howevtv, 



this decrease has reached a minimum, and from this 

 point the gas behaves in the ordinary way, and the product 

 pv increases with the pressure in accordance with the 

 equation given above. Not the least sign of liquefac- 

 tion occurs at this temperature at any of the pressures 

 employed. 



This is an observation of great importance. It shows 

 that at sufficiently low temperatures hydrogen behaves 

 on compression like all other gases, and has a minimum 

 value of pv occurring, not at some exceptionally low 

 pressure, but at a pressure of fourteen atmospheres. 



For the general representation of the four isothermals 



U) I) XI 1(1 rn 



Meter 



In the sketch given of the apparatus used when the gas was cooled bj- means of liquid oxygen, a is the bulb in which the compression of the hydrogen takes 

 place. The vessel containing this is filled with oxygen under pressure and surrounded by ethylene under evaporation, liquid ethylene bein^ 

 obtained by passing the gas already cooled by a mixture of ice and salt through a worm tube immersed in r in a mixture of solid carbon dioxide 

 and ether. By suddenly diminishing the pressure on the oxygen by opening the tap v communicating with the outside air, the gas is partially 

 liquefied and the desired temperature is reached. The volume of the gas in a is afierwards measured by allowing it to escape into the eudiometer E. 



thus obtained, Wroblewski employs an empirical formula 

 that differs but slightly from one given by Clausius— 



-b = — — - ^ 



the values of the constants being in the present case R 

 _ _i^^.^ a = o'ooiii665,K =0-00051017, and e = 1-003892. 

 This equation serves for the calculation of the critical 

 temperature, pressure, and volume of hydrogen, for which 

 the values are found to be— 



. , 8K RT , , 



6e^ = — ^s-» "■ = o-j and0 = 30, 

 ayRa 8a ' ^ ' 



or, substituting the values for the constants, we get- 

 Critical temperature 6 = 32-6 or - 240° '4 C, 

 ,, pressure ir — 13*3 atmospheres, 

 ,, volume (p = 0-00335, 



from which 



Critical density = o'02'j. 



Similar results were obtained by employing other formulae 

 than the one given above, but this was retained as being 

 in best agreement with the experimental results. 



These numbers afford at once an explanation of the 

 ordinary behaviour of hydrogen on compression and the 



