40 Mr. J. F. Daniell on Crystallization. 



represented in the last figure. If, from some predisposition of 

 affinity, the particles of any solution should continue to com- 

 bine in this definite proportion, a number of cubes would be 

 formed ; which, again attracting one another, would unite toge- 

 ther side by side, according to the general laws which we have 

 observed. A compound cube would be thus naturally con- 

 structed, and it is evident that mechanical force would resolve 

 such a solid into a number of smaller cubes ; for upon the 

 planes of junction the spheres of one cube are only held to the 

 spheres of another cube by the binary attraction of two par- 

 ticles for each other, while in every other direction each ball 

 is in contact with three others at least. 



Now, the first simple cube may be resolved into two similar, 

 but irregular tetrahedrons ; and if we suppose an octohedron 



formed by a decrement upon the angles of the cube analogous 

 to that in M. Haiiy's cubic system, these tetrahedrons would 

 exactly fill the intersticial spaces ; and an octohedral arrange- 

 ment would be formed with precisely the same angles as the 

 constructions which we have previously considered, but which 

 would exhibit the phenomena of the cubic instead of the octo- 

 hedral cleavage. 



