AS CLASSIFICATION CHARACTERS IN MINERALOGY. 437 



essential -espeet as a mineral, to iron pyrites, than to these latter sub- 

 Stances. It is utterly impossible not to admit this. Nevertheless, if 

 we blindly follow the chemical view, we are actually forced to maintain 

 the contrary. In the well-known Krystallo-chemiische System of 

 Gustav Rose, for example — a system held up by many as a perfect 

 model — magnetic and copper pyrites are not only widely separated 

 from iron pyrites, but they are placed in the same general division 

 with the red silvers, zinkenite, jamesonite, &c. And, in like manner, 

 the carbonates and titaniates, the silicates and sulphates, &c, stand 

 together, from the assumed constitution of their respective acids. 

 Arrangements of this kind may be to a certain extent convenient, but 

 who will venture to call them anything more. Natural classifications 

 akin to those of the botanist and zoologist, most assuredly they are 

 not. Although opposed to my earlier belief, I now feel confident that 

 a satisfactory classification of minerals will never be accomplished 

 until the mineralogist cease altogether the attempt to force his 

 groupings into corelation with the present views of Chemistry. Let it 

 not be forgotten, that Mineralogy has in more than one instance, when 

 in seeming opposition to Chemistry, led the chemist to the adoption of 

 new principles by which the harmony of the two sciences has been 

 maintained ; and hence it may be legitimately inferred that, if the 

 mineralogist proceed fearlessly to classify the objects of his study 

 without regard to the restrictions which Chemistry would set before 

 him, further means of agreement will be found to reconcile any differ- 

 ences that may spring up from this independent method of procedure. 

 At present, Chemistry is to the mineralogist, in many respects, a tyrant 

 the most absolute, compelling him by its exactions to groupings in 

 which natural analogies have not the slightest voice. If two com- 

 pounds have the same representative formulae, or if amongst binary 

 compounds of oxygen or sulphur, for example, the basic elements 

 happen to be isomorphous or otherwise related in the simple state* 

 they must be placed in the same group, no matter how loudly their 

 physical characters and general conditions of occurrence may exclaim 

 against it. In this manner, in a mineral classification beyond com- 

 parison the most philosophical in its general features yet arrived at, 

 we have the unavoidable union of carbonic acid gas with sassolin (hy- 

 drated boracic acid) and quartz : the three occurring together, as 

 binary oxygen compounds, the respective bases of which (carbon, 

 boron, silicon,) happen in the simple state to be of a kindred nature. 

 In the system of Gustav Rose again, arsenic acid and iron-glance are 

 placed in the same group, simply because the two are sesqui-oxygen 

 compounds ; a collocation permissible, perhaps, in the case of sesqui- 



