iMINEEALOCY; WiTIl A CLASSIFICATION OF SlljlCATKS. 39 



2. The couceptiou that the kws of progressive or homologons series, previously recog- 

 nized, only in hydrooarbonaeeous bodies, must be extended to mineral species, and are of 

 imiversal application. 



3. The conception that the variations observed in the chemical composition of such 

 mineral species are due, not only to their highly polybasic character, but also, in certain 

 cases, to indefinite admixtures of homœomorphous species, as previously indicated by 

 Delesse, Scheerer and Von Waltershausen, extended and generalized by myself, and subse- 

 quently adopted by Tschermak. 



4. The attempt to fix the molecular weights of such compounds as the polysilicates 

 and polycarbonates from their densities as compared v\àth those of species, the minimum 

 molecular weights of which are otherwise determined ; and the assumption that for 

 homœomorphous solids, and probably for all, the molecular volumes are identical. 



5. The adoption of atomic formulas to represent the composition of mineral species, 

 and the showing that comparisons of the volumes or spatial relations of complex species 

 like the silicates, should be based on the numbers which are deduced from these atomic 

 formulas, and which represent the relative volumes of the unit -weight in the species com- 

 X^ared. P being the unit-weight got by dividing the empirical molecular weight by the 

 number of oxyd-atoms in the formula (including any chlorid, fluorid or sulphid atoms 

 which may be present), and D the specific gravity, the volume of the viuit, designated as 

 V, is represented by the quotient got by dividing P by D. 



6. The showing that, in related and homologous species, the hardness and chemical 

 indifference are inversely as the value of V; or in other words, tliat they increase with the 

 condensation, the relative amount of which in the species compared is shown by the 

 diminution of V. 



III. — A Classification of Silicates. 



§ 36. In order to show the application of the principles already set forth in the second 

 part of this paper to the classification of mineral species, I have chosen the Natural Silicates, 

 for the reason that, while they present the most complex and the largest group of native 

 species, their physical and their chemical history alike have been more thoroughly studied 

 than those of any other group. To this it may be added that the writer has, for many 

 years, employed a classification based on these principles for the arrangement of his own 

 private collection of the native silicates. These maybe regarded as constituting one great 

 natural order, and, withoirt again adverting to the importance of considering in a natrrral 

 system alike the chemical and the physical history of species, it may be said that a funda- 

 mental distinction in silicates is that presented by their chemical constitution, as including 

 either protoxyd or sesquioxyd bases, or both united ; for which reason the order Silicate is 

 divided into three suborders : Protosilicate, Protojiersilicate, and Persilicate. 



The names of protoxyd and protosalt for ferrous compounds, and of peroxyd and per- 

 salt, instead of sesquioxyd and sesquisalt, for ferric compounds, are familiar to chemists; 

 and when, in naming the suborders of silicates it became necessary to select a term to 

 designate alike ferric compoiands, aluminic compounds, and those in which ferric and 

 aluminic sesquioxyds partially replace each other, I ventured to substitute for sesquisilicate 

 and protosesquisilicate the shorter and more euphonious names of persilicate and proto- 



