122 



EEPORT — 1886. 



lated it at '044. This is, therefore, a tolerably satisfactory agreement. 

 The two curves meet at about 0-3°. 



For benzene W. Fischer found +5-8° as melting-point; Regnault had 

 found +4-35°. The equation for the vapour-pressure over solid benzene 

 was found to be 



^=24-985 -I- 1-6856^-^0-031339^2 ; 



that for the vapour-pressure over liquid benzene was 

 _^=26-40 + l-429.5)' + 0-04505i2. 



The two curves do not meet at 5"3°, but they should meet on some point on 

 the line of solid-liquid, and may do so at some point corresponding to a 

 pressure higher than atmospheric. From the diagram it appears that the 

 two curves would not meet for some distance from the melting-point; 

 this is not certain, but it is so probable as to point to some error, perhaps 

 arising from impurity of the benzene. 



Yapour-pressures of Mercury. 



The determination of the relations between temperature and vapour- 

 pressure for mercury was found by Regnault to be very difficult, on 

 account of the occurrence of violent bumping ; the results are published 

 in ' Memoires,' t. xxvi. p. 520. The temperatures were measured by an 

 air-thermometer of constant volume, but of small initial pressure, it 

 having been partially exhausted before the beginning of a series of ex- 

 periments. The formula log Y^a + ha' + cif was used, where F is the 

 vapour-pressure, and the constants determined by a sufficient number of 

 data from observations including a wide range ; thus the table on pp. 

 -520, 521 was calculated from the formula. Regnault's observations were 

 too few and too doubtful, and the results given by him for vapour-pres- 

 sures of mercury at low temperatures and at ordinary temperatures have 

 been proved to be quite illusory. 



A few other physicists have attacked this question ; among them 

 Hagen, McLeod, Hertz, and Drs. Ramsay and Young. 



We will first give Hagen's results,' comparing some of them with 

 Regnault's : — 



Hagen's differ -widely from Regnault's numbers, but, unfortunately, 

 there is good reason for thinking Hagen's results entirely untrustworthy 

 in spite of the very great care which he took to avoid sources of error. 



An experiment of McLeod's made the early numbers of Hagen (those 



' Wiedemann's Annalen der PhysiTt und Chemie, 1882, xvi. p. 610. 



