2 NOTE ON STELLIFORM CRYSTALS. 
exhibit rounded and confluent outlines; whilst those produced by 
crystallization, are made up of plane surfaces, meeting, in sharp edges, 
under definite (and for the same substance, under constant) angles.* 
Although crystals usually originate when matter passes slowly from 
the gaseous or liquid condition into the solid state, crystallization and 
solidification are not actually identical. Various substances, for 
example, such as silica in certain conditions—its hydrate (constituting 
the different opals)—gums,—certain resins, &c.,—appear to resist 
altogether the action of crystallization. Mr. Graham (the present 
Master of the British Mint) has suggested that these bodies may 
retain, or retain to a greater extent than crystalline bodies, the latent 
heat which they possessed before solidification. 
The crystal forms and combinations met with in Nature, exclusive of 
those produced by the chemist in his laboratory, are exceedingly 
numerous, many thousands being known to exist. By the help of 
certain laws, however, and, more especially, by the aid of one, termed 
“the Law of Symmetry,”’ we are enabled to resolve these multitudinous 
combinations into six groups or systems. The forms of the same 
group combine together, and may be deduced mathematically from 
each other; whilst those of distinct groups are unrelated. Thus, 
although the cube, the rhombic dodecahedron, and the regular 
octahedron, appear at first sight to be unconnected forms, yet by the 
Law of Symmetry their co-relations may be readily shown. This 
law, for instance, exacts one of three things, of which the most 
important is to this effect, viz., that if an edge or angle of a crystal 
be modified in any way, all the similar edges or angles in the crystal 
must be modified in a similar manner. Now the cube has twelve 
similar edges and eight similar angles. Consequently, if one edge or 
one angle be truncated, or, to use a term more in conformity with 
the aciual operations of Nature, if one of these be suppressed during 
the formation of the crystal, all the other edges (or angles) must be sup- 
pressed equally ; and if the new planes which thus arise be extended 
until they meet, the rhombic dodecahedron on the one hand, and the 
* This law is affected within slight limits by isomorphous replacements, and also by 
changes of temperature. The law itself appears to have been discovered by Nicolaus Steno 
(then a naturalized Florentine) as early as 1669, but its true importance was not appreciated 
until the re-announcement, or rather re-discovery of the law in 1772 by the French erys- 
tallographer, Romé de V’Isle. Many of the contemporaries of the latter—amongst others, 
the celebrated Buffon—attempted to deny its existence, but being susceptible of practical 
proof, its truth was soon established. 
