520 



Mr. J. B. Hannay. 



[Mar. 10, 



of heat and the more the production of sonorous vibrations, for a 

 greater number of air atoms will collide with warmed surfaces in a 

 given time. That this assumption is justified is proved by the fact 

 that this absorption of radiant energy renders the particles of disso- 

 ciated nitrite of amyl and other vapours visible as well as warm, and 

 therefore they can assume dimensions that are comparable with the 

 particles of smoke. 



II. " On the Limit of the Liquid State." By J. B. Hannay, 

 F.R.S.E. Communicated by Professor G. G. Stokes, 

 Sec. U.S. Received February 22, 1881. 



(Abstract.) 



In this paper the author gives an extended account of the work under- 

 taken to determine whether the liquid state extends above the critical 

 temperature, or whether it is bounded by an isothermal line passing 

 through the critical point, as had been indicated in a former paper. 

 A large apparatus was constructed, with several improvements before 

 described, details of construction being given in the full paper. It 

 was found that manometers with small bores gave higher readings 

 than those with larger internal diameter, so the manometers used were 

 of the largest size compatible with the strength required to resist the 

 pressure. The thermometers were carefully prepared by heating and 

 cooling, and compared with the standard at Kew. All the usual pre- 

 cautions were taken to obtain accurate numbers. The critical tem- 

 perature and pressure of the liquids were first determined accurately, 

 and then a quantity of a gas insoluble in the liquid was compressed 

 over the liquid, and the critical temperature again determined under 

 increased pressure. When the densities of the two bodies, e.g., 

 alcohol and hydrogen, are far apart from each other, the gas shows 

 no effect in lowering the critical temperature, as is the case with 

 carbon dioxide and air, whose densities approach much nearer, but 

 simply acts as a spring against which the upper surface of the alcohol 

 bears, thus having a surface free for observation at pressures far 

 above that of the vapour of the liquid. When the liquid passes the 

 critical temperature at any pressure the meniscus is lost, and the fluid 

 freely diffuses into the superincumbent gas, but this does not occur at 

 temperatures below the critical, except where very high pressure has 

 made the gas appreciably soluble in the liquid. Thus the curve of 

 vapour tension, that is, the curve representing the temperatures at 

 which a given pressure will produce liquefaction suddenly becomes 

 isothermal at the critical point, and passes along the co-ordinate 

 denoting the critical temperature. 



As surface tension is the only property by which the liquid state 



