11 



species, two, or even three minerals, may be pre- 

 sent in nearly equal proportions. Moreover, in 

 these compounds, as it is not unfrequent to find 

 the component minerals alternately predominant, 

 it must follow that different varieties of the same 

 rock would be found in different divisions, and, 

 of course, under different aames. We need not 

 look far for examples of confusion under such a 

 system : granite offers a very obvious one. 



Such are the inconveniences that must follow 

 from adopting the presence of any particular mi- 

 neral as the principle of arrangement. But the 

 same reasoning applies to any attempt to arrange 

 rocks according to the degree of their composition, 

 or the number of substances of which they are 

 compounded. 



If they are to be divided into binary or ternary 

 compounds, it is easy to perceive that the same 

 effect of separating those which are naturally con- 

 joined, even in the same specimen, and of uniting 

 those which are entirely distinct in every other 

 essential character, will follow. Thus, for exam- 

 ple, the granites of three ingredients would be 



