44 DE. THOMAS STEEEY HUNT ON THE 



ratio (excluding a little water) of 4:9, not far from that of pectolite, with which 

 we have placed them. Different analyses haA^e assigned to talc the ratios for the fixed 

 basis of 2 : 5 and 1 : 3, (the water being variable), — the latter corresponding to sepiolite, 

 1 : 3 : 1. For neither of these do we know any corresponding pectolitic silicate. 



§ 85- We come, in the last place, to the quadrisilicates, for which we have no repre- 

 sentatives in the table among anhydrous or among hydrous magnesian species. They are, 

 however, represented in the pectolitic group by no less than three species, okenite, 

 apophyllite, and an unnamed species got artificially by Daubrée. It is fibrous like okenite, 

 is decomposed by acids, and is a hydrous silicate of lime, with six per cent, of soda, giving 

 the ratios, 1 : 4 : |-. Pectolite, it will be recollected, contains in like manner about nine 

 per cent, of soda, while apophyllite contains five per cent, of potash and a little fluorine. 



§ 86. The process by which this vxnnamed pectolitic silicate was obtained by Daubrée 

 is very instructive, as showing, in many ways, the action of heated water on an undifieren- 

 tiated silicate of igneous origin. He took for the subject of his experiments a common 

 glass, the analysis of which gave silica G8.4, alumina 4.9, lime 12.0, magnesia 0.5, and soda 

 14. Y = 100.5. Tubes of this glass were sealed up, with many precautions, in tubes of iron, 

 with about one third their weight of pure water, and exposed during several weeks to a 

 temperature not less than 400° 0. At the end of this time the glass was found to be com- 

 pletely disaggregated and changed into a white fibrous or lamellar substance, composed in 

 great part of the fusible pectolitic quadrisilicate of lime and soda in question. "With this 

 were found abundant crystals of quartz, and a few crystals haAuug the form of diopside, 

 and the composition of a lime-iron pyroxene. In certain of the crystals of this latter mineral 

 were also included microscopic grains of a black matter resembling magnetite or picotite, 

 probably the former. The iron of these minerals was perhaps derived from the metal tube. 



§ . 8*7. The net result of the prolonged action of heated water on the glass was that the 

 vitreous silicate gave up 44.0 per cent, of its silica, 64.0 per cent, of its soda, and 85.0 per 

 cent, of its alumina ; the lime, with the remaining silica and soda and alumina (equal to 1.4 

 hundredths) forming the pectolitic silicate. Of the separated silica, the larger part sei^ar- 

 ated in the form of well-crystallized quartz, with globules of chalcedony, and the few crys- 

 tals of pyroxene mentioned aboA^e. The soluble matter, got by treating the decomposed 

 glass with boiling water, was a silicate of soda, with some dissolved alumina, neglect- 

 ing which the proportions of soda and silica in the liquid were found, in one instance, to 

 be as 63 : 3*7 by weight, corresponding to an oxygen-ratio of R : Si of aboiit 3 : 4. But as, 

 according to Daubrée's analysis, 85.0 per cent, of the alumina had passed into the solu- 

 tion, this would make for 63 parts of soda not less than 9. "I parts of alumina, which 

 should give for the silico-aluminate in sohition a ratio of E. : r : Si of nearly 3 : 1 : 4 ; a 

 result of much significance which it would be A'ery desirable to A^erify by further trials. 



§ 88. Daubrée has recorded experiments like that aboA^e made to determine the 

 solA'ent action of heated water upon Aatreous volcanic rocks, such as obsidian and perlite, 

 which gaA^e similar result to glass, though, according to him, not so well defined. Frag- 

 ments of sanidin, of oligoclase, of potash-mica and of pyroxene, in these tubes, suffered no 

 apparent change, though incrusted with crystals of quartz deriA^ed from the glass. This 

 stability was to have been expected from the fact that crystals of pyroxene are formed 

 under similar conditions, and, as we shall see, both albite and orthoclase have since been 

 crystallized at high temperatures in presence of solutions of alkaline silicates. Another 



