•384' Prof. Norton on Molecular Physics. 



be admitted that solidification is in every instance but a more or 

 less perfect crystallization, it will be perceived that the investi- 

 gation of the mechanical process of crystallization must consist 

 essentially in an inquiry into the conditions and results of the 

 operation of the molecular forces under special circumstances. 

 The general nature of these forces, and the laws of the variation 

 of effective molecular action with the distance between two mo- 

 lecules, have already been under discussion. We have also seen 

 (pp. 382, 383) that the mechanical condition of an individual 

 molecule is subject to change under the operation of heat, and 

 external forces generally, by reason of a change produced either 

 in the dimensions or form of the atmosphere of the molecule, 

 and that it may thus acquire permanent axes of attractive force. 

 To establish a sufficient basis for a general explanation of crys- 

 tallization, we have only to remark further that the molecule of 

 every particular substance has primarily and inherently its own 

 special physical condition, by virtue of which it exercises an 

 effective action that would be represented by a special curve, and 

 experiences under the operation of heat and other causes its own 

 peculiar changes of mechanical condition. Upon this idea it 

 may be seen that every substance may have its particular form 

 of crystallization, although the molecules should be devoid of all 

 natural polarity. 



The different systems of crystallization may be regarded as so 

 many different systems of equilibrium of masses of molecules, 

 under the operation of molecular forces diversely modified by 

 the circumstances that determine the crystallization. The gene- 

 ral nature of the modifications consists in a spheroidal form im- 

 parted to the molecular atmospheres, and the consequent deve- 

 lopment of certain axes of attraction — that is, of diameters of 

 least or greatest length, in the direction of which the attraction 

 has at a given distance a maximum or minimum value, and the 

 limit of stable equilibrium (Qa, fig. 1, p. 203) a minimum or 

 maximum value. 



Crystallization begins at a certain point of a liquid, and is 

 generally determined by the loss of heat, or the evaporation of 

 the liquid solvent. We already have seen reason to believe that 

 the molecules of the liquid have a symmetrical arrangemeut 

 previous to the crystallization (p. 279). Whether this be ad- 

 mitted or not, such an arrangement obtains in the crystal formed 

 from the liquid. The particles successively take positions in the 

 corners or angular points of a series of polyhedral figures ; as 

 cubes, prisms, octahedrons, &c. Any two such figures lying 

 contiguous to each other, have a common face, or, as in the case 

 of the octahedron, a common angular point. The crystalliza- 

 tion takes place either successively or simultaneously in the 



