VII.] CRYSTALLIZATION. 75 



forms, having unlike sets of faces, but within the limits of 

 the same species, (that is to say, having the same crystal- 

 lographic elements,) so that the two halves of the crystal 

 seem as if they belonged to two different crystals. Hemi- 

 hedrism and hemimorphism are characteristic of particular 

 species. 



Hemimorphism, which is occasional among crystals, is 

 all but universal among organisms. Among plants and 

 sponges there is a difference between the root and the 

 upper part ; and among animals generally, between the 

 two ends of the alimentary canal. 



Crystalline grow^th is controlled in a very remark' Crystals 

 able way by the laws of symmetry. Crystals have aj'^P?"^^^" 

 power comparable, or rather identical, at least in its effects, 

 with the power of organisms to repair injuries. Lavalle 

 has found^ that if a crystal of alum is placed in pure 

 water till its edges and angles are dissolved away, and 

 then put into a solution of alum, the edges and angles 

 wiU form again. He also found that if one of the angles 

 is cut off such a crystal, and the crystal put into the 

 solution lying on the cut-off angle, so as to prevent that 

 angle from forming again, a corresponding truncation will 

 form, as the crystal grows, on the opposite angle- 

 Crystalline formation is also dependent in a very Fortiis of 

 remarkable way on the medium in which it takes place : ^"^^*^g^^^^''' 

 and I intend farther on to show reason for believing that by the me- 

 the same is true of some of the lower tribes of organisms, f^."™ rmed 

 Beudant has found that common salt crystallizing from in. 

 pure water forms cubes, but if the water contains a little 

 boracic acid the angles of the cubes are truncated. And 

 the Eev. E. Craig has found that carbonate of copper, 

 crystallizing from a solution containing sulphuric acid, 

 forms hexagonal tabular prisms; but if a little ammonia 

 is added, the form changes to that of a long rectangular 

 prism with secondary planes on the angles. If a little 

 more ammonia is added, several varieties of rhombic 

 octahedra appear ; if a little nitric acid is added, the 



1 My authority for the statements in this and the following paragraph 

 is Dana's Mineralogy, vol. i. p. 138. 



