468 Mr. W. Sutherland on the 



the substance, and R, the usual constant for it when a perfect 

 gas, T the absolute temperature. Thanks to Amagat's in- 

 valuable supply of data (Ann. de Ch. et de Phys. [6] xxix.), 



I have been able to find an equation of more extensive range, 

 but as this is still in manuscript, I shall use here the pub- 

 lished equation referred to. In that paper, p. 276, values of 

 M 2 / for water are given, M being the molecular mass which 

 for H 2 is 18. With the dyne as unit of force these are 



II X 10 12 from the latent heat of vaporization, 9 x 10 12 from 

 the critical data, and 6 x 10 12 from the capillary data. The 

 last value was obtained on the assumption that in water the 

 molecules consist of (H 2 0) 2 , but M in M 2 Z is only 18. As 

 the latent heat of vaporization of water includes also the heat 

 of dissociation of its complex molecules, and as the critical 

 pressure probably has its value seriously affected by dissocia- 

 tion, the value 6 x 10 12 must be the most nearly correct, as I 

 bave hitherto always treated it to be. Then changing to the 

 atmo as the unit of pressure, when RT at 0° C. has the value 

 11200/9, we get for (H 2 0) 2 or 1 at 0° 



v"T |^=12790, 



but as 



df dv ldv ldv _ ldv Idf df 



dy~~dTldf' ■•'" vdf~ vdTJdT~ Po ! p dT 



(17) 



using the value of k x , namely "0009, we have for the calcu- 

 lated compressibility of 1 at 0° the value '000016. Similarly 

 at 50° we get '0000235 . For higher temperatures it seems 

 to me safest to proceed thus. According to van der Waals's 

 principle of correspondence, the compressibilities of 1 and of 

 ethyl oxide at low pressures will always be in the same ratio 

 to one another if taken for comparison at temperatures which 

 are equal fractions of their absolute critical temperatures. 

 We may take the absolute critical temperature of 1 as the 

 same as that of water, namely 641, and that of ethyl oxide 

 as 468. In the next table are given certain temperatures C 

 for 1 and the temperatures C for ethyl oxide which correspond 

 to these ; then the compressibilities of ethyl oxide at these 

 temperatures, and finally the compressibilities of 1 (dihydrol) 

 calculated from those for ethyl oxide on the principle that 

 the ratio at all the corresponding temperatures is that which 

 holds for 1 at 50° and ethyl oxide at —37°, obtained by 

 extrapolation from Amagat's data from 0° to 200° between 

 50 and 100 atmos. 



