184 
molecules of different substance, as indicated by tbe different 
markings. The crystal is commencing from the central cube; 
to which, according to probability, the others add themselves 
two by two or otherwise according to the laws ot 
loqenesis , which we are beginning slowly to understand. The 
point essential to my argument is the simplicity of the process 
of crystallization, resembling the addition of brick to brick by 
the builder ; the completed wall, however extensive, being the 
result of a thousandfold repetition of the same act. 
85. The next figure that I shall place under the eye of the 
reader represents the probable formation of a rhomboidal crystal, 
commencing by the assemblage of two molecules ( a a ) oscillating 
a 
a 
round their axis, till, the opposite poles approaching, they fix 
themselves in the position of greatest proximity indicated, the 
other molecules (b b) adjoining themselves subsequently at a 
greater distance. , , 
36. The rough copy I have made does not attempt to do 
iustice to the beautiful drawing of the original; winch indicates 
the horizontal plan of four molecules of epidote, and (by an 
arrangement of M. Gaudin's) denotes the constituent atoms ; 
neither can I give the mathematical calculations, for all which 
I refer to the original work. The point which I insist upon is 
the simplicity of the nature of crystallization, and the apparent 
analogy of the forces which bind together atoms into molecules 
* Mrnide des Atomes, p. 66. 
