146 Royal Society : — 



at the time being j and thus, by its tendency to expand, this liquid 

 and anything enclosed in the tube were subjected to a very great pres- 

 sure. By keeping the tubes in various parts of the house, according 

 as the weather varied, I have been able to maintain for several 

 weeks or even months a pressure of, for instance, about 100 atmo- 

 spheres, as measured by means of a capillary-tube pressure-gauge 

 enclosed within the larger tube. Since in all cases I had a second 

 tube which from first to last was treated precisely like the other, 

 pressure excepted, I have been able to determine the effect produced 

 by the pressure with very considerable accuracy — at all events so as 

 to leave no doubt whatever about the general facts. At the same 

 time I wish it to be understood that the results described below must 

 be looked upon only as approximations to the truth. 



I will first call attention to the well-known influence of pressure on 

 the fusing-point of various substances, since it is a connecting link 

 between well-established facts and those I am about to describe. 

 Bunsen* and Hopkins j* have shown that substances which expand 

 when fused have their point of fusion raised by mechanical pressure ; 

 that is to say, since mechanical force must be overcome in melting, 

 the tendency to melt must be increased by heat before that opposition 

 can be overcome ; and the pressure required to keep them solid at 

 any temperature above their natural point of fusion may be looked 

 upon as the mechanical representative of the force with which they 

 tend to fuse at that temperature. Professor W. Thomson % has 

 shown that, on the contrary, water, which expands in freezing, has 

 its point of fusion lowered by pressure ; that is to say, since mecha- 

 nical force must be overcome in crystallizing, crystallization will not 

 take place under increased pressure unless the force of crystalline po- 

 larity be increased by reducing the temperature. Thus, calculating 

 from his experiments and from the known latent heat of ice, and 

 assuming that no heat is gained or lost by contact with external objects, 

 if we had 1 part of ice and 100 of water at 0° C, and then applied 

 a pressure of 103 atmospheres, the ice would, as it were, dissolve in the 

 water, the whole would become liquid, and the temperature be reduced 

 to —'792° C; or, in other terms, at that temperature the tendency 

 to crystallize is exactly counterbalanced by that pressure. 



Now I find that similar principles hold true with respect to the 

 solubility of salts in water. If, when they dissolve, the total bulk 

 increases, pressure reduces their solubility ; whereas if the bulk de- 

 creases, pressure makes them more soluble ; in other words, solution 

 or crystallization is impeded by pressure according as mechanical 

 force must be overcome in dissolving or in crystallizing. 



Various authors have written on the volume with which salts enter 

 into solution § ; but since the subject before us requires a different 



* Pogg. Ann. 1850, vol. lxxxi. p. 562. 



f British Association Report, 1854, p. 57. 



X Trans. Roy. Soc. Edinb. vol. xvi. p. 575. 



§ Playfair, Chem. Soc. Quart. Journ. vol. i. p. 139. Michel and Krafft, Ann. 

 de Chitn. 3 ser. vol. xli. p. 471. Schiff, Ann. der Chemie, vol. cix. p. 325 ; vol. cxi. 

 p. 68 ; vol. cxiii. p. 349. Gerlach, Specifische Gewichte der Salzlosungen, &c, 



