( 40fi ) 



secretion shows itself, justifies the supposition thai the amphibole 

 must be said to belong to the ferruginous varieties, and from this 

 may be concluded thai the ratio of the oxyde-quanta MgO and FeO- 

 which after the crystallization of the amphibole remained available 

 for the formation of rhombic pyroxene could not be altered very 

 much at the expense of MgO, compared with the ratio which the 

 norm-calculation yields. These difFerenl grounds, therefore, tell against 

 hypersthene, and for bronzite or ferruginous enstatite. In accordance 

 with our conclusion, as will appear further on, unmistakable enstatite- 

 (bronzite-)crystals of greater dimension- are found in the glassy 

 dacite 11 591» of the Soengei Embahoe, which has a (dose chemical 

 and mineralogie similarity with this rock. There, too, the glass- 

 base possesses the same characteristic rhombic pyroxene microlites, 

 whose bronzite-nature is made probable by the chemical analysis. 



Fe.,< ). ( being reduced to FeO, the molecular proportions, calculated 

 on a sum of 1<>(). yield the following values (Tin added to SiO s ): 



Sin, Al<), FeO Mg<> C;K) Na 8 K„0 

 75.45 10.27 2.72 L.56 4.22 4.I55 1.43 



from which follows the formula according to Osann: 



s A C V a c f n m k Series 



I 75.45 5.78 4.4i» 4.02 8.09 6.28 5.62 7.5:5 10.00 1.4(5 « «• 



11 75.45 5.78 4.22 4.29 8.09 5.90 (5.00 



III _ 12.14 9.42 8.43 



IV _ — (12.14 8.85 9.00 



These figures require some explanation. As appeared already at 

 the calculation of the norm, there remained after saturation of the 

 alkalis and (JaO with Al a 0, a not unimportant quantity of sesquioxyde. 

 The analyzed material being absolutely fresh and even microscopi- 

 cally free from even trace of decomposition, there is no possibility 

 of a relative increase of ALU., by a removal of alkali- and OaO : 

 a consequent distribution of the Al,0 8 -remainder over alkalis and 

 OaO 1 ) could not be reasonably defended here. Also the dissolution 

 In the andesite-magma of fragments from an adjacent rock rich in 

 aluminium cannot bul be highly improbable; both macroscopically 

 and microscopically there is not the leasl indication for it. For the 

 rest the ALO.-remainder cannot be found in amphiboles or pyroxenes 2 ) 



i| I,. Milch, Ueber Spaltungsvorgange in granitischen Magmen nach Beobach 

 tungen ini Granit des Riesengebirges, Festscbrift H. Eosenbusch, 1906, 127. 



~) A. Osann, Versuch einer chemiscben Classification der Ei uplivgesteine, Min 

 Petr. Mill. XXU, 1903, ool . 



