5o6 



NA TURE 



[March 29, 1894 



THE BEHAVIOUR OF LIQUIDS UNDER 

 HIGH PRESSURES. 



ONE of the most important generalisations which has 

 been obtained in recent years from the study of 

 the effect of tempf^rature and pressure on the volume of 

 stable liquids and gases may be expressed by the law 

 that if the volume of a given mass of substance be kept 

 constant, increase of pressure is proportional to increase 

 of temperature. This relationship was proposed as early 

 as 1878 by Levy, who was indeed anticipated to some 

 extent by Dupre in 1869, but was first set upon a firm 

 experimental basis, at least for vapours, by Ramsay and 

 Young in 1887. They represent it algebraically by the ' 

 equation p-bt-a, in which p is the pressure, / the 

 temperature, and b and a are constants which vary with I 

 the volume and the chemical nature of the substances 1 

 employed, and the curve corresponding to this equation 

 they term an isochor. The law may therefore be shortly 

 expressed by stating that for stable substances the | 

 isochors are straight lines. This generalisation leads, 

 as Fitzgerald has shown, to the significant conclusions 

 that specific heat at constant volume must be a function 

 of the temperature only, and internal energy and 

 entropy must be expressible as the sum of two functions, 

 one of which is a function of the temperature only, and 

 the other a function of the volume only. 



The experiments of Ramsay and Young extend at 

 most over a pressure range of about i 00 atmospheres, 

 and it thus becomes a matter of considerable interest to 

 ascertain if the linear isochor still persists under pres- 

 sures which are very much higher, especially when the 

 substances operated upon are liquids at temperatures 

 which are well below their critical temperatures. Im- 

 portant data on this point may be gleaned from Xos. 92 

 and 96 of the Bulletin of the United States Geological 

 Survey, wherein are grouped together accounts of the 

 varied researches carried out during the last few years by 

 Mr. CarlBarus on several of the high pressure phenomena 

 of liquid substances 



His earlier work (5?^//^/z'« 92), completed in 1889, dealt 

 with the isothermal compressibility of some fourteen 

 liquids at temperatures and pressures having values as 

 high as 310' and 600 atm. respectively — the pressure 

 range being thus six times as great as that employed by 

 the English observers. 



From the data obtained, isochors^ were eventually 

 deduced with the result that, although below 180" they 

 pursued a linear course, above this temperature under the 

 high pressures employed they gave definite indication of 

 being curved. To test by careful experiment over still 

 wider ranges of pressure, this important question of cur- 

 vature is the object of the later observations of Mr. 

 Barus, which are detailed in No. 96 of the Bulletin. 



The principle of the method there described consists 

 in keeping the volume of the substance constant, and 

 directly measuring the pressures which it supports at 

 different temperatures, and thus obtaining immediately 

 the data necessary for plotting the isochors. A new 

 compression pump was devised, by means of which the 

 enormous pressures of 2000 or 3000 atm. could be exerted. 

 The temperature range was similar to that of the previous 

 experiments, and the substances operated upon were 

 ether, alcohol, thymol, pura-toluidine, and diphenyl- 

 amine. 



The results thus obtained pointed conclusively to the 

 fact that at high temperatures and high pressures the 

 isochors of the liquids employed are really curved. In 

 general the linear isochor persisted up to pressures of 

 1000 atm., and over temperature ranges which varied 

 with the nature of the substance, the maximum 

 temperature being about 115' in the case of ether, and 



1 Instead of isochor, Mr. Barus uses throughout his papers the term 

 isometric, originally proposed by Willard Gibbs. 



65" in the case of diphenylamine for the volumes used. 

 These volumes, it may be mentioned, were not measured, 

 so that no stress can be put upon the absolute slopes 

 obtained for the curves. In all cases but that of thymol, 

 the deviation from the straight curve was a marked 

 abrupt phenomenon, and occurred generally between 

 1000 and 1500 atm. Thymol, however, gave no appre- 

 ciable deviation. 



What this curvature may mean is as yet but a matter 

 for speculation. The author inclines to the belief that it 

 indicates a change of molecular state, and evidence into 

 which he does not enter may be taken to point to the 

 same conclusion. Ramsay and Young's work on dis- 

 sociating gaseous substances like nitrogen peroxide and 

 acetic acid has shown that during the progress of mole- 

 cular decomposition curved isochors are obtained, which 

 apparently bridge over the gap between the linear 

 isochors corresponding with the simple and complex 

 molecular states. If a like explanation applies to the 

 liquids at present under consideration, it leads to rather 

 a curious result, for the curvature of the isochor for 

 alcohol is in the opposite direction to that of the isochors 

 for all the other liquids. It would therefore follow that 

 under increasing pressure at constant volume the altera- 

 tion of the molecular state of alcohol is in the opposite 

 sense to that of all the other liquids, and the observed 

 direction of curvature favours the view that the liquid 

 alcohol molecule eventually becomes simpler, while those 

 of the other liquids become more complex. Such a con- 

 dition of things may the more readily be conceived when 

 it is borne in mind that the general physical behaviour 

 of alcohol, and more especially its behaviour with 

 regard to surface energy, indicate that under ordi- 

 nary conditions it probably contains molecular aggre- 

 gates, the complexity of which alters as the tempera- 

 ture alters. A liquid like ether, on the other hand, 

 seems to contain under ordinary conditions simple 

 gaseous molecules. The above results would thus have 

 the interpretation that, volume remaining constant, the 

 complex molecule of liquid alcohol corresponding with 

 the origin of the isochor remains of the same degree of 

 complexity over wide ranges of pressure until in the 

 region of high pressures itbecomes less complex ; whereas- 

 the simple molecules of a liquid like ether, under the 

 same conditions, eventually become associated into more 

 complex aggregations. Of course, until more data have 

 been accumulated, the above explanation must be re- 

 garded as but a conjecture ; indeed, any definite reason 

 why the molecular complexities of liquids like alcohol and 

 ether should be so different, under ordinary conditions,. 

 is at present entirely wanting. Whatever happens,, 

 the significant observations here considered have defi- 

 nitely shown that the law of linear isochors, although it 

 is valid throughout wide variations in the external con- 

 ditions, eventually breaks down in the region of high 

 pressures and high temperatures. 



No less striking results are obtained by Mr. Barus on 

 the effect of pressure in solidifying a liquid. Here he 

 studies the volume changes produced by pressure during 

 the solidification and fusion of naphthalene at various- 

 constant temperatures, and he is thus enabled to plot 

 several of the isothermal lines for liquid-solid naphthalene 

 between the temperatures of 60" and 130', and between- 

 the pressures of 40 and 1700 atm. 



The remarkable result arrived at in this way is, that 

 during change of state the " on" curve obtained by in- 

 creasing the pressure and passing from liquid to solid is 

 quite distinct from the " off'" curve obtained on passing 

 from solid to liquid. At any temperature 'solidification 

 always occurred at a higher pressure than that at which 

 the solid fused. This is, of course, in harmony with the 

 well-known fact that the temperature of the ordinary 

 melting point of a substance is in general higher than, 

 its temperature of solidification. 



NO. 1274, VOL. 49] 



