327 



lute tho'jgh important characters decide tlie groups aul families ; and 

 lastly, slight external differences, if nothing better rem lins, serve to 

 distinguish between closely allied species. 



It is not necessary, in a general view, such as that taken in this 

 paper, to give the details of the classification. Bat the outlines of the 

 plan will convey you a clearer idea of what is new in the method. 



Silicates are so numerous ani play such an importiut paH in the 

 mineral kingdom lint they may be said to constitute one sub-kingdom, 

 all other mioerals being releo-ated to form another. Thus we have the 

 two fundamental divisions : — the Non-Silicates and the Silicates. 



In the sub-kingdom of the Non-Silicates the classification is easy 

 whatever system is adopted. We may either adopt the classification 

 bsiSei upon the acidic element or the one which depen Is upon the 

 basic or metallic constituent of the compound. 



In any case we shouli have, in the beginning, a table containing 

 the analytical tests which characterise, first, the great division of 

 Silicates ; secondly, each group of the Silicate or Non-Silicate sub-king- 

 dom. At the head of eich group, the characteristic reactioa of each 

 family would be given ; and fiially, the chiracf"er3 neglected till then 

 wjuld point oat each single species which shoul.l now hi completely 

 described. 



In the division of the Non-Silicates you perceive that in most 

 cases the analytical test leads the student infallibly to the group and 

 the family, and very often to the very species ; leaving veiy little to 

 the physical and crystalline character in the work of determination. 



Among the Silicites, however, the difficulty is greiter. The 

 division into Hydrous al^ I Anhydrous cannot well be effncted in every 

 case by analysis. Furthermore, the composition, so far as the quantity 

 of each element is oncerned, h extremely vaiiible, Here, therefore, 

 according to the plan we have adopted, we must follow, for the deter- 

 mination, the method which has served to constitute the species. 

 Disregarding then, in the work of determination, those theoretical 

 groups base 1 on the similarity of chemical formulas, we may proceed 

 with the help, firstly, of chemical analysis whenever it is practically 

 available; secondly, the crystalline characters, including cleavage; 

 thirdly, those physical properties which are more intimately connected 



