38 Mr, J. F. Daniell on Crystallization. 



of which it was found that crystalline bodies may be cloven in 

 certain determinate and constant directions, from which the 

 planes of least resistance in the solid are easily determined ; 

 but this is not the only force by which their structure may be 

 dissected. The lower degrees of chemical affinity may be 

 applied, as I have formerly shown, (Journal of the Royal 

 Institution, vol. i. p. 24,) more delicately to the same purpose. 



An irregular mass of alum, which, although to the eye it 

 exhibits no traces of crystalline arrangement, may easily be 

 shown to possess as regular a structure as the best defined 

 crystal of the same substance. Mechanical force will not avail 

 us for this purpose, as regular cleavages cannot be detected 

 in it; but if we expose it to the solvent power of water, 

 at first the fluid acts upon the salt with so much energy as to 

 overcome the cohesion of the solid in every direction alike ; 

 but, as the water becomes saturated, its power diminishes, and 

 it is nearly balanced by that delicate modification of cohesion 

 upon which crystalline structure depends. The consequence 

 is, that the solid now yields to the solvent only in the points of 

 least resistance, and the mass will present the form of octo- 

 hedrons and sections of octohedrons, as it were, carved or 

 stamped upon its surface. 



The numerous forms which are thus dissected from the mass 

 are arranged in a definite order, with regard to each other and 

 the different faces of the mass ; and the series which occur 

 upon one face, and those which correspond with it, are never 

 intermingled upon dissimilar faces. Thus in one direction the 

 light will be reflected from the faces of octohedrons and sec- 

 tions of octohedrons all upon the same plane ; and by turning 

 the mass upon its axis, the same will be repeated at every 

 quadrant of a circle. By gently inclining the mass, the re- 

 flection will next arise from right-angled parallelograms of 

 every dimension, which are similarly repeated upon turning the 

 mass upon its axis. 



Now, by supposing the process of solution continued till the 

 several planes intersected each other, it is clear that various 

 modifications of the octohedron and cube would result, all 

 necessarily referable to the same structure of particles in the 

 original mass : and it is obvious that each of the almost infinite 



