September 30, iSy;] 



NATURE 



535 



I 



axis, decreases from the centre of the field both in the direction 

 of increasing and decreasing pressures. The tenour of these re- 

 sults is an additional indication of the recurrence of a lower 

 critical temperature at which cycles must necessarily vanish. 

 The decrease of the breadth of the cycles in the direction of de- 

 creasing pressures suggests the possible occurrence of a point in 

 the region of negative external pressure, so circumstanced that 

 beyond it the substance would solidify at a lower pressure than 

 that at which it fuses. This may be interpreted as follows : 

 the normal type of fusion changes continuously into the ice- 

 type of fusion through a transitional type characterised by the 

 absence of volume lag. 



An independent discussion, more searching in character, has 

 quite recently been given by Tammann. Tammann points out 

 that for the normal case of fusion and for increasing pressure, 

 the two determinative factors of the Clapeyron equation — the 

 volumes and latent heat of fusion — will not in general simul- 

 taneously become and remain zero. He argues that the volume 

 constant will at the outset decrease with pressure passing through 

 zero to negative values. Hence the curve representing the rela- 

 tion of melting-point to pressure must initially rise to a maximum 

 when the melting-point pressure ratio is zero, and then decrease. 

 Contemporaneously the latent heat of fusion, decreasing con- 

 tinually with pressure, eventually also reaches zero, but at a 

 much later stage than the volume constant. At this stage, i 

 therefore, since melting-point and the volume constant now have j 

 definite values (the latter negative), the melting-point and pres- 

 sure ratio is negatively infinite. Hence the curve expressing 

 the relation of melting-point to pressure decreases with increasing 

 pressure from the maximum specified as far as the pressure at 

 which latent heat is zero, and there drops vertically downwards. 

 Thus Tammann's melting-point pressure curve, with its initial 

 and final ordinate in the direction of temperature, maps out a 

 field of pressure and temperature, within which the body 

 is solid. Outside of this region the body is liquid, and 

 cannot by pressure alone be conceivably converted into 

 the solid state. Any thermodynamic change involving a 

 march through the boundary of this region is accom- 

 panied by the discontinuity of fusion, of viscosity, &c. A 

 march through the final ordinate (for which latent heat is zero) 

 is probably not accompanied by such discontinuity. For a 

 given temperature there may be two fusion pressures. At a 

 temperature sufficiently below the melting-point, the continued 

 increase of pressure should therefore move the normally fusing 

 body from the solid into the liquid state continuously. This is 

 a somewhat anomalous result of close reasoning ; but it must not 

 be forgotten that in the depth of our ignorance of the actual 

 occurrences above several thousand atmospheres, the term 

 anomaly is a misnomer. Indeed, if we regard the melting- 

 pressure curve beyond the stated maximum as characterising 

 the ice-type of fusion (which Tammann does not do), our diffi- 

 culties would in a measure be reconciled. 



Tammann finally points out that the term lower critical 

 temperature is not justified by the character of the phe- 

 nomenon. Data for melting-point and pressure, due to 

 Damien, seem directly to corroborate the occurrence of 

 zero values in the ratio of melting-point and pressure in- 

 crements, but Damien's tests are restricted to a pressure in- 

 terval much 100 small to be trustworthy. Of the two bodies 

 which have been tested throughout long-pressure intervals, 

 naphthalene shows a linear melting-point and pressure ratio for 

 2000 atmospheres, while the carbon tetrachloride of Amagat, 

 though initially concave upwards, soon also becomes linear. 

 Clarence King has therefore, in geological considerations, so re- 

 presented it. To conform with Tammann's inferences the 

 interior of the earth would have to be a fluid. 



One point of issue, however, in these cases is clear : at 

 Andrews' critical temperature both the difference of specific 

 volumes and the latent heat of fusion vanish simultaneously 

 wherever observed. Under corresponding conditions of change 

 from liquid to solid, the internal pressures are of tremendously 

 greater value for both states, and the passage of the solid into 

 the liquid molecule may involve an immense transfer of energy 

 without any corresponding change of volume : for the density 

 of the molecule itself eludes observation. The manner of this 

 isothermal change from one state to the next is in all cases along 

 the characteristic doubly inflected contour first pointed out by 

 Thomson for vapours, and since elaborated by Van der Waals, 

 Clausius and others. We may for brevity call this a voUtvie lag, 

 and measure it in terms of the pressure or the volume interval 



NO. 1457, VOL. 56] 



subtended. The liquid can exist even above the critical tempera- 

 ture, which would mean that even here pressuremust be reduced 

 below the critical pressure in order to rupture the liquid molecule. 

 Pronounced as these phenomena are for the change from gas 

 to liquid, they become much more remarkable, indeed often 

 formidable, for the change from liquid to solid. In this case a 

 volume lag subtending more than lOO 'atmospheres is the rule ; 

 in other words, it takes a much greater pres.sure to solidify a 

 liquid at a given temperature than to liquefy the solid. Among 

 all these cases there is a group of well-known bodies in which, 

 while the solidification pressure is of marked intensity, the iso- 

 thermal pressure of spontaneous fusion may even be below zero, 

 or be in the region of negative pressure. Take the single 

 example of thymol, among many : this body between zero 

 Centigrade and its melting point at 53°, can be kept in either 

 the solid or the liquid state at pleasure. Given at about 50" 

 in the liquid state it would require more than 2000 atmospheres 

 to solidify it. If solid, it must obviously remain so even if 

 pressure be wholly removed. But thymol may be similarly 

 treated, beginning with the under-cooled liquid state at 28°, i.e. 

 25" below its melting point. Even here at least one thousand 

 atmospheres are needed to condense it (400 have been tried quite 

 ineffectually). Once solid, it would require about 1000 atmo- 

 spheres of negative external pressure again to melt it. In other 

 words, it could not be melted again on the same isothermal. 



If we but knew more about the physical constants involved in 

 these transformations, we could predict the results along the 

 lines of J. W. Gibbs's splendid theory of the equilibrium of 

 heterogeneous mixtures ; but with the dearth of our concrete 

 knowledge of long range physical phenomena relating to liquids 

 and solids, we must be content with humbler methods. 



I have always regarded the significant behaviour, instanced for 

 the case of thymol, as capable of a broad interpretation. Profs. 

 J. J. Thomson and Fitzgerald in the British Isles, and Elihu 

 Thomson in America, have recently sought for atomic dissocia- 

 tion in the electrolysed vacuum of a Crookes' tube. Speaking 

 to the same point, I would venture to assert that we may 

 reasonably look to the volume lag for a rational account 

 of the genesis of atoms. We have already met with 

 two orders of volume lag : the first at the mergence 

 of gas into liquid being usually a few atmospheres in iso- 

 thermal value ; the second at the mergence of liquids into solid, a 

 hundred or even one thousand times as large in isothermal value, 

 and characterised by the fact that, whereas freezing pressures 

 may be enormous, the corresponding isothermal melting pres- 

 sure may even be markedly negative. 



If then we further inquire as to what will happen if we in- 

 definitely compress the solid along a suitable isothermal, I think 

 it is logically presumable that, with the succeeding and pro- 

 foundly accentuated volume lag, we shall reach the next atom 

 in a scale of increasing atomic weights. 



However enormous the condensation pressure for this purpose 

 may be, it is supposable, in the light of the examples already 

 given that, along an accessible isothermal, the disintegrating 

 external pressure of the new atom may be permanently negative. 

 Hence the new atom will persist within the pressure and 

 temperature range available in the laboratory. 



But the last stage is virtually identical with the first, or the 

 inherent nature of these changes is periodic. The inference is 

 therefore that, under suitable thermal conditions and continually 

 increasing pressure, the evolution of atoms, of molecules, of 

 changes of physical state, again of atoms and so on indefinitely, 

 are successive stages of periodically recurring volume lag. 



UNIVERSITY AND EDUCATIONAL 

 INTELLIGENCE. 



The Rev. A. M. Fairbairn, Principal of Mansfield College, 

 Oxford, will distribute the prizes to the successful students of 

 St. Thomas's Hospital on Saturday, October 2, at 3 p.m., in 

 the Governors' Hall. 



Prok. Robert G. Aitken and Mr. W. H. Wright have 

 been appointed assistant astronomers at the Lick Observatory. 



A NEW centre of the London County Council Technical 

 Education Classes will be opened next Monday evening, 

 October 4, at the Charterhouse and Rogers' Memorial Institute, 

 Goswell Road, E.C. The committee of the school announce 

 that workmen's classes in workshop arithmetic, workshop 

 chemistry and physics, workshop drawing and mechanics will 

 be conducted. 



