44 



SCIENCE PROGRESS 



we find that 



p-po _ R 



p 



T 



= — — ^- and hence 



As 



a con- 



T-T v T """ - T 



sequence, if the perfect gas state can be assumed at very small 

 pressures and densities, the absolute value of zero C. is given 

 by the inverse of the coefficient of expansion from o*o° C. 

 to some temperature which is not only ioo° C. most suitably 

 for the reason given above, but also because this is the standard 

 interval of the centigrade scale. 



In the following table are collected the values for a few 

 gases used for thermometric purposes in which the value of B is 

 known to a sufficiently high degree of accuracy to make the 

 calculation of any real value. 



The curvature of the isotherms is so small that the C term 

 does not enter into the result except for the purpose of obtaining 

 the theoretical normal volume. 



If the critical data were known with sufficient accuracy, it 

 would be possible to derive these results by substitution in the 

 reduced form, but at present the errors are far too great to make 

 this method of any real value. 



Table I. Absolute Zero 



According to Berthelot, who has carefully reviewed the 

 whole of the data available, the most probable value for absolute 

 zero is — 273-09° C, while the above results for hydrogen and 

 helium, which were obtained subsequently, give 273*1. Thus it 

 is probable that the uncertainty has now been reduced to a 

 hundredth of a degree centigrade. 



There is a much greater" degree of uncertainty in the 

 evaluation of the divergences of the gas scales from the absolute, 

 if the values given by different workers are given an equal 

 weight. However, it is most probable that the values calculated 



