CRYSTALS. 283 



another a regular octahedron, and a third a dodecahe- 

 dron ; forms corresponding to those of the preceding line. 



In view of its simplicity, the regular system may be 

 regarded as a sort of primitive architecture, yielding, 

 however, to no other system in the beauty of its forms. 

 Under one or the other of these systems all forms of 

 crystals are included. To each of them, with the ex- 

 ception of the last, belong innumerable sets of forms 

 according to the degree of inequality or inclination of 

 the axes. Equality and rectangular position of the axes 

 being characteristic of the first system, it is not suscep- 

 tible of the sort of variation which is essential to pro- 

 duce different sets of figures. But in this, as in other 

 systems, the modification of surfaces may occur to any 

 extent. 

 , , ,, 707. As the architect is able, from some 



bn.ow now trie 



formofacrys- relic of a broken column, to build up in 

 ferr^ifrom, imagination the temple of which it formed 

 its angles. a p art . as the comparative anatomist knows 

 how, from the fragment of a single bone to reconstruct 

 in imagination the perfect animal which possessed it ; 

 so, from the merest point of a crystal, its complete form 

 may often be readily inferred. In proportion as a dou- 

 ble pyramid is lengthened out, the angles above arid 

 below are rendered more acute. Prom an accurate 

 admeasurement of this angle its whole shape may there- 

 fore be inferred. Such admeasurement of various an- 



