ON THE SOLUBILITY OF SALTS. 297 



the one crystallizing in the form of rhombic, and the other in quadratic 

 tables or Hat prisms. He has likewise obtained* a salt isomeric with 

 common Epsom salt, MgO, SQ 3 + 7HO, and more soluble than it; it cry- 

 stallizes in rhomboidal tables, which lose their transparency when taken out 

 of their mother-liquor. Marignae has shown f that the quadratic crystals of 

 sulphate of nickel have the formula NiO, S0 3 + 6II0, and that an isomeric 

 salt crystallizing in the monoelinic system is formed at a temperature of 

 from 50° to 70° Cent. Further researches will make us acquainted with 

 many other examples of the same kind. The salts of manganese, cobalt, 

 chrome alum, &c, present us with remarkable examples of molecular changes 

 accompanied by striking changes of colour, in some of which we have true 

 isomeric salts, as in the case of manganese, and in the case of chrome alum 

 a difference in the amount of hydrated water. We should, however, distin- 

 guish clearly between the molecular changes which give rise to the isomeric 

 salts, and those which are accompanied by a total alteration of physical pro- 

 perties. It is worthy of remark, too, that many chemists have been led, from 

 wholly different reasons, to assume a more complex (as to the number of 

 molecules) composition for salts and for chemical compounds generally. I 

 may mention, lor example, Avogadro's view, that the equivalent of a sub- 

 stance does not necessarily represent the relative weight of an integral mole- 

 cule, but that the latter may be a multiple or an aliquot part of the former ; 

 and Hunt's!, that all solid substances which have the same form, contain 

 an equal number of atoms in equal volumes, a view which leads him to pro- 

 pose such a formula as Na,„ Cl 10 for common salt. The existence of double 

 salts of isomorphous bases containing a great number of equivalents of one 

 combined with one or more of the other, and which are easily decomposed by 

 merely dissolving them in a large quantity of water, is another circumstance 

 which renders the hypothesis of Favre and Silbermann, and the application of 

 it in a modified form above given, of increased interest. 



If we adopt the hypothesis that crystallized salts are isomeric compounds, 

 and that solutions are unstable compounds of salt molecules and water 

 molecules, it seems to me that the phenomena of crystallization and solution, 

 and perhaps the whole phenomenon of the change of physical state of bodies, 

 may be referred to the category of double decomposition. It would lead me 

 too far to attempt to develope this hypothesis here ; indeed, without a good 

 deal of experimental data to sustain my arguments, it would not be useful. 

 I hope to be able to develope it fully on another occasion. I have mentioned 

 it now, merely because it is intimately connected with the remaining 

 questions which 1 have set down above, and which, with the exception of the 

 fourth, I do not propose to discuss on this occasion. 



There would obviously be a wide distinction between a solution containing 

 the molecules MgO, S0 3 +7HO, and one containing only MgO, S0 3 . When 

 we dissolve Epsom salt in water, which of the two solutions have we ? The 

 answer to this question depends to a considerable extent upon what may be 

 the result of inquiries concerning the nature of solution. If a solution be a 

 chemical combination between salt molecules and n molecules of water, the 

 compound being liquid, and furthermore, if the crystallized salt be an 

 isomeric form of the salt in solution, we must conclude that the water of 

 the hydrated salt, as such, does not exist in combination with the salt mole- 

 cule in solution, but becomes attached to it at the moment of the formation 

 of the crystalline salt by a process of double decomposition. In assuming, 



* Ann. de Chim. et de Phys., vol. xliii. p. 405. 

 ■)■ Liebig and Kopp's Jahresbericht, 1855, p. 411. 

 J Silliwan's American Journal (2), vol. xv. p. 116. 



