MINKIIALÔGY; WITH A CLASSIFICATION OF SILICATES. 79 



higli temperatures iinder pressure, is changed into crystalline qiiartz, pyroxene and a 

 pectolitic silicate, or when at ordinary temperatures gelatinous precipitates become crys- 

 talline in an aqueoTis medium. In like manner, the amorphous persilicatcs left by the 

 decay of aluminous spathoids such as feldspars, pass into phylloid species like kaolinite 

 and pyrophylite, or into adnmnntoid species like andalusite, librolite and cyanite, while the 

 spontaneous change of certain anhydrous colloids into the crystalline state at ordinary 

 temperatures is seen in the well known examples of vitreous arsenic trioxyd and barley- 

 sugar. The above considerations as to the nature and relations of colloids, upon which the 

 writer has elsewhere insisted, will help to show their importance in a scientific study of 

 mineralogy.' 



The same mineral types which serve to divide each of the suborders of natural 

 silicates into well-defined tribes reappear in the non-silicated oxyds, and serve for their 

 classification. Reserving for another occasion the details of classification of this great 

 order of OxYDATES, Ave may note that while the Oxyadamantoid tribe embraces such 

 species as periclasite, chrysoberyl, the spinels, magnetite, corundum, diaspore, hematite, 

 quartz, rutile, cassiterite, etc. ; the Oxyspathoids include cuprite, zincite, creduerite, pyro- 

 lusite, tridymite and senarmontite ; and the Hydroxyspathoids, gibbsite, gothite, and man- 

 ganite. Among the Oxyphylloids are brucite, pyrochroite, massicot, minium, melaconite, 

 hydrotalcite and pyraurite ; while the Oxjcolloids or Opaloids embrace bauxite, limouite, 

 opal, uran-gummite and eliasite. 



§ 105. The plan of the present essay does not embrace a discussion of the species of this 

 order, but it will be advantageous in connection with the history of the silicates, to notice 

 some facts regarding the atomic volume of certain of these oxyds. The adamantoid tribe 

 of the Oxydâtes includes a large number of species crystallizing alike in the isometric and 

 the rhombohedral systems, which give for V a value approximating to that of the adaman- 

 toid silicates, chrysolite, pyroxene, garnet, epidote, beryl and tourmaline. C'h. Grcrhardt, 

 in 1847, pnblished a note on "The Atomic Volume of some Oxyds of the Regular System," 

 which was translated and given in English by the present writer in the same year." There- 

 in accei^ting the view held by Laurent (§ 27) of an indefinite or unlimited divisibility of 

 the molecule, G-erhardt proposed to reduce to a common foimula M.O, (m^o, of our 

 present notation) not only protoxyds like periclasite, and protoperoxyds like the spinels, 

 magnetite, chromite and franklinite, but sesquioxyds like martite, hematite, and braunite, 

 and titanates like perofskite and mcnaccanite, including thus not only isometric and 

 rhombohedral, but tetragonal forms. In the fractional formulas then proposed by him 

 were employed the atomic values for aluminium, ferricum, chromicum and titanium, 

 which have since been adopted by the writer (§ 68). For all of the species above named, 

 reduced to his common formula of M/), Grerhardt deduced atomic volumes varying from 

 10.6 to 11.4, which, divided by two, to correspond with our chemi('al irnit, give for the 

 value of Y from 5.30 to 5.70. These variations were ascribed by Gerhardt to errors in the 

 determinations of density, to impurities, and to the difficulty of taking into account small 

 portions of various oxyds present, and he conceived that for all of these species the atomic 



' See, for a further discussion of this subject, tlie writer on the Origin of CrystalHnc Kock.s ; Trans. Roy. Soc. 

 Canada, Vol. ii. Sec. ?>. pp. 55-57. 



' Ann. de Pharmacie, 1847, xii, .'!81-3S5, and Ainer. Jour. Science, 1847, iv. 40,5-408 ; .see al.so Jl>li?., isr)2, xiii. 

 370-372, where tlie sanic. subject w;U3 further ihsciissed by the present writer. 



