182 Prof. Sydney Young on the 



of the volumes of liquid and of saturated vapour to the tem- 

 perature or pressure, as shown in the accompanying diagrams 

 constructed from the results with ethyl alcohol. 



A very small alteration of temperature, such as 0*1°, at or 

 just below the critical point, produces a considerable alteration 

 in the volume ; therefore in order to obtain a direct reading of 

 the critical volume it is necessary that the substance shall be 

 exactly at its critical temperature. 



I have assumed that a substance is in this state when on 

 rapidly increasing the volume somewhat above the critical 

 volume the fall of temperature due to the expansion causes 

 a momentary separation of liquid and vapour. 



In order to determine the critical volume, I note the position 

 of this temporary mark of division and then diminish the 

 volume slightly. After waiting a few minutes for the tem- 

 perature to become constant again, I increase the volume very 

 slightly but rapidly, and again note the position of the mark 

 of division, which is now nearer to the top of the tube. 

 Proceeding in this way it is possible under favourable con- 

 ditions to make the substance occupy such a volume that a 

 very slight but rapid expansion gives a temporary mark of 

 division of liquid and vapour almost exactly at the top of the 

 tube. This volume I take to be the critical volume, and I 

 have succeeded in determining it directly in the case of 

 benzene, fluorbenzene, and acetic acid ; while with chloro- 

 benzene, for which mercury vapour was employed as a jacket, 

 it was only possible to obtain a rough approximation to the 

 true volume on account of the slight unsteadiness of the 

 temperature. 



The critical volumes of the other substances may probably 

 be ascertained with fair accuracy in the following manner : — 



At low temperatures and pressures the ratios of the mole- 

 cular volumes of liquid and saturated vapour of any one 

 substance to those of fluorbenzene at corresponding tem- 

 peratures and pressures differ somewhat widely as a rule, but 

 as the critical point is approached the differences diminish 

 and at the critical point itself all four values should, of course, 

 be identical. It follows, therefore, that by mapping the ratios 

 against temperature, four straight lines or curves should be 

 obtained, which, when produced, should cut one another at the 

 critical temperature, and the point of intersection should give 

 the ratio of the molecular critical volume to that of fluorbenzene. 



It will be seen from the accompanying diagrams, figs. 3 to 6, 

 Plates II. & III., that in the case of benzene and acetic acid the 

 four curves do very nearly cut one another at the critical 

 temperature, and the ratios of the critical volumes thus 



