-1859] WRITINGS OF JOSEPH HENRY. 109 



saltpetre and common salt — be dissolved together in a suf- 

 ficient quantity of water, and the liquid be suffered gradu- 

 ally to evaporate, they will be found at the bottom of the 

 vessel in separate crystals. The cubes of common salt can 

 readily be distinguished from the long-sided prisms of salt- 

 petre, and when these are chemically analyzed, each is found 

 to be exclusively composed of its respective substance. Not 

 a single atom of the saltpetre is found in the crystal of salt, 

 nor one of the latter in the former. The same effect takes 

 place if magnesia and saltpetre be dissolved in hot water and 

 the solution be suffered to cool. The case however is al- 

 together different when sulphate of magnesia, and sulphate 

 of nickel or sulphate of zinc are crystallized together, from 

 the same solution. The separation of the two substances 

 does not take place as in the former instance; the individual 

 crystals formed will contain both sulphate of zinc and sul- 

 phate of magnesia, or sulphate of nickel and sulphate of 

 magnesia, and this in every possible proportion, according 

 to the relative amounts of the two salts in solution. Now if 

 we compare a crystal of sulphate of magnesia with a crystal 

 of sulphate of nickel, we find they have identically the same 

 crystalline form : there is no perceptible difference in their 

 angles, edges, or solid angles. And since a large crystal is 

 built up of an aggregation of small ones of the same form, 

 it is evident that the primitive molecule of sulphate of nickel 

 must have the same form as that of the sulphate of mag- 

 nesia; and therefore that in forming a large crystal they 

 may be mingled together in the way we have just described, 

 provided they are of the same size, or perhaps some multiple 

 of the same size, for it is evident that it would be impossible 

 to build a wall of symmetrical structure with bricks of dif- 

 ferent angular forms and sizes, since the parts would not fit 

 or exactly fill the spaces. We must therefore conclude that 

 though the ultimate atoms of bodies may be spherical, the 

 groupings of them, which form the primitive crystallizing 

 molecules, are of different geometrical shapes and sizes. 



The atomic weights or combining proportions. — Though the 

 primordial atoms may all be of the same weight and size. 



