Chemical Equilibrium of Solids. 27 



can be explained away, direct experiments on metals confirm- 

 ing Osmond's view will not be easily forthcoming. Two 

 years ago I spent much time in endeavouring to throw light 

 on the polymerization of metals, by studying the hydroelectrics 

 of stretched and soft wires*, by dissolving hard-drawn and 

 annealed iron, by investigating the carburation changes of 

 strained steel ; but in no case have I found evidence of an 

 indisputable bearing on the point at issue. Finally, regarding 

 the mere statement f of inferences of the above kind, I believe 

 I anticipated Osmond by interpreting viscosity in solids | 

 with reference to Clansius's theory of electrolysis. 



2o. The proposition of § 23, taken in connexion with my 

 observation of the occurrence of volume-lag % under isother- 

 mal conditions, suggests important inferences on the behaviour 

 of stressed solids generally, and leads naturally to an explana- 

 tion of hysteresis. If the pressure under which a liquid just 

 solidifies, cost, par., is in excess of the (positive or negative) 

 external pressure at which it again liquifies (volume-lag §), 

 and if the chemical equilibrium of the solid acted on changes 

 with each change of strain, then, quite apart from considera- 

 tions directly involving the second law of thermodynamics, 

 quite apart from the energy dissipated in the cycle, the solid 



* American Chemical Journal, xii. p. 1 (1890). 



t American Journal, xxxiii. p. 28, January 1887. 



\ American Journal, xxxviii. p. 408 (1889). 



§ A good example of the volume-lag is given by an undercooled liquid 

 like fused thymol (melting at 53°), for instance, which at, say, 30° solidities 

 under about 500 atm., and does not again liquefy on removing the ex- 

 ternal pressure. I have worked with para-toluidine, naphthalene, a-naphtol, 

 paraffin, and caprinic, pa]mitic, and monochloracetic acids. My original 

 belief in the general character of these static phenomena has recently 

 been disturbed by new experiments, in which I found for the special case 

 of paraffin that it would be possible to be misled by local solidifications 

 of the column under pressure. The volume-lag is always marked when 

 there is even a suggestion of undercooling. The action of pressure is 

 then similar to the closing of a door and opening it again without lifting 

 the latch. Observations made at intervals of 20 minutes each, with 

 palmitic acid, showed that at 66° this substance solidities between 300 

 and 400 atm., and then remains solid even between 200 and 300 atm. 

 At 100°, a-naphtol solidifies at 500 to 600 atm., and when pressure is 

 being removed it remains solid until about 150 atm. Undercooled 

 caprinic acid at 20° solidifies under 150 atm., remaining solid. At 33° 

 it solidifies at about 500 atm., and then fuses again at about 200 atm. 

 Local solidifications here suggest the occurrence of isomers or other 

 impurity. Sometimes, as in the case of naphthalene, the volume-lag 

 is not thoroughly static, but becomes viscous in character, &c. 1 have 

 yet to study the effect of temperature on the magnitude of the volume- 

 lag, i.e. on the difference of pressures isothermally corresponding to 

 solidification and to fusion. Very large pressure-intervals are essential 

 for this purpose. 



