and the actual Forms (rf' Inorgank Bodies, 137 



ship, from which it follows, that that form must be the most 

 symmetrical, which has the greatest number of parts similar in 

 relationship in a given space. Guided hy such an idea, in deter- 

 mining what form is the most symmetrical, we are at once neces- 

 sitated to fix upon the regular polyha?dron, whose facets are all 

 similar to each other, equally distant from the centre, and so mi- 

 nute as to be single particles. For, in such a figure, the rela- 

 tionships of all the parts are identical, and, at the same time, 

 the number of such parts in a given space is a maximum. Of 

 all possible forms, therefore, the regular polyhaedron, whose fa- 

 cets are single particles, is the most symmetrical. But such a 

 polyhaedron is just a sphere. The sphere, therefore, is the limit 

 towards which bodies, while their symmetry is becoming more 

 perfect, must tend. And all bodies, did the particular forms and 

 forces of their particles not present obstructions in the way of 

 such a result, and were they aggregated into individuals, ac- 

 cording to the law of greatest symmetry alone, wholly unmodi- 

 fied, must have been contained under spherical contours. All 

 this results from our idea of symmetry. Let us see, then, what 

 are the forms which the power of crystallization, so far as we 

 can obtain evidence, tends to develope. And here it must be 

 remarked, that, so long as we are not at liberty to assume any 

 thing as to the forms of the crystallizing molecules of mineral 

 bodies (and certainly nothing could justify such assumptions, as 

 that they are all cubes, or spheres, or spheroids, or identical 

 with the cleavage forms, or the hke), we must just make use of 

 such knowledge of the structure of crystalline bodies as we pos- 

 sess, and of such observations, as their actual forms, and the 

 phenomena of their increment and decrement, display. 



Proceeding in this way, we find that about a fourth part of 

 the whole crystalline series, possess such a structure, that their 

 cleavage forms are tessular, or such as may be inscribed in a 

 sphere ; and what is especially to be remarked of such crystals, 

 is the fact, that their actual forms are never less, and tmiaily 

 more nearly spherical than their cleavage forms. Thus the dia- 

 mond has an octohedral cleavage form, and its actual form is not 

 only never more dissimilar to a sphere t' an the octohedron, but 

 often, by replacements and truncations, is made to approximate 

 the sphere to such a degree, that it more nearly resembles that 



