RICHARDS AND MAKE. — THERMAL EXPANSION OF HYDROGEN. 137 



The old result of Regnault,* 0.003G613, is indeed not very far from the 

 new one, but has too large a possibility of error to afford important 

 evidence. More recently, the coefficient of expansion at constant pressure 

 has been studied very carefully by Chappuis, f who finds that the volume- 

 change per degree under constant pressure is 0.0036600, a value notice- 

 ably lower than ours. The pressure was, however, different from ours, 

 and the range of temperature greater. 



The value corresponding to tension-increase in constant volume has 

 been found by three separate experimenters recently to be about 

 0.00366256,J as follows : 



Chappuis 0.00366256 



Travers 0.00366255 



Onnes 0.00366257 



0.00366256 



There can be no doubt that the value for constant pressure is less than 

 this, because j> v with hydrogen undoubtedly increases as the volume 

 diminishes. According to Regnault, if pv at a pressure of one meter of 

 mercury is taken as 1.00000, that at 2 meters is 1.00043, that at 4 meters 

 is 1.00153, and that at 8 meters is 1.00424. By plotting these figures as 

 ordinates and the pressures as abscissae, there is obtained, as is well 

 known, a curve which is almost linear, and is therefore capable of accu- 

 rate use in graphic calculation. The application of this 1 curve to the pres- 

 ent case is now to be discussed. 



Imagine a mass of gas at the temperature 32.38° under atmospheric 

 pressure compressed further until it assumes the volume occupied by the 

 same mass of gas at 0° and atmospheric pressure. Its tension will now 

 be 1.12 atmospheres. For this increased tension, by a slight extrapola- 

 tion of Regnault's curve, it is easy to find that the value for p v would be 



0.0044 

 about 1.000044§ or = 0.037 per cent of the increase in pressure. 



Thus it would appear that the pressure-coefficient in constant volume 

 should be about 0.037 per cent greater than the volume coefficient under 



* Regnault, Memoirs de 1'Acade'mie rles Sciences, 21, 116, 426 (1847). 



t Chappuis, Rapports du Congres International de Physique, 1, 133 (1900). 

 Trav. et Mem du Bur. Int. Poids et Mes., 13 (1903.) See also Travers, The Ex- 

 perimental Study of (iases, p. 149 (Macmillan, 1901). 



t Travers, Chem. News., 86, 61 (1001 ?). 



§ Various graphic calculations gave figures varying from 1.000012 to 1.000046. 



