a7id the actual Forma of fnorgank Bodies. 139 



is the limit of perfect symmetry. In imperfect crystals also, in 

 which the summit only is complete, that summit will generally 

 be found to display similar phenomena, indicating an approxi- 

 mation to a hemispherical contour. To conclude, let the reader 

 inspect the 197 crystalline species, of which representations and 

 descriptions are to be found in Mohs' Mineralogy, and in the 

 whole series he will only find five or six which are destitute of 

 those features of approximation towards the sphere of which I 

 speak, while in all the others, the principle now advanced will 

 be found most palpably displayed. 



• Nor is the fact to be viewed as in any degree adverse to the 

 view now advocated, that many implanted crystals occur in 

 prisms, whose length is frequently many times their diameters. 

 For, in consequence of the manner in which, there is reason to 

 believe, that implanted crystals are evolved, the prismatic form 

 must necessarily be generated, or the process of evolution must 

 cease. But on this subject I do not enter. Meantime it may 

 be remarked, that the transverse striae upon crystals of quartz, 

 when viewed with a magnifier, serve well to illustrate the deter- 

 mination of crystals towards spheroidal forms, even when cir- 

 cumstances limit them to prisms. 



During the increment and decrement of crystals also, well 

 marked phenomena, pointing to the same principles, may be 

 detected. Thus, when a cube of rock-salt is exposed to a 

 damp atmosphere, from which it attracts moisture, the dissolu- 

 tion begins regularly at the edges, so that each of the original 

 edges is speedily replaced by two planes. Then these gradually 

 increase till the hexahedron is transformed into a trigonal icosi- 

 tetraedon, a six-sided into a twenty-four-sided figure, of very re- 

 gular and symmetrical aspect, and very much more similar to the 

 sphere than the original cube. (Mohs, vol. ii. p. 38). 



In the external forms of crystals, when viewed with common 

 light, however, but a small part of the spheroidal features and 

 tendencies which their structure possesses are visible. ' The 

 view now advanced is far more fully exhibited in crystalline 

 bodies, when polarized light is applied to them, — nay, in ligiit 

 itself, whose phenomena may be called ethereal crystallizations, 

 and wherein, as well as in that department of nature which we 

 have just examined, symmetrical arrangements, and a cotiattis 



