CHEMICOMINERAL RELA TI ON SHI PS IN ROCKS 235 



orthosilicate and liberate silica, increasing the amount of quartz 

 or raising ferromagnesian orthosilicates to metasilicates. But 

 the development of biotite would draw upon the ferromagnesian 

 orthosilicate (Mg,Fe) SiO s , reducing the amount that might be 

 converted into metasilicate. The quartz limit would then be 

 shifted toward lower silica percentages. In orthoclase-leucite 

 rocks, or leucite rocks without orthoclase, the development of 

 this potash-ferroaluminous molecule would reduce the amount 

 of these minerals and liberate silica, which would raise the 

 metasilicate leucite to orthoclase. The effect would thus be 

 to cause the leucite limit to recede toward lower silica per- 

 centage. 



It has already been noted that the development of biotite 

 involves, besides the potash-bearing molecule under discussion, 

 an orthosilicate of iron and magnesium, which is the same as an 

 olivine molecule. The development of this molecule in rocks that 

 might otherwise carry ferromagnesian metasilicate molecules 

 would still further liberate silica, to appear as more quartz or 

 raise more leucite to orthoclase. 



In case alumina is in excess of potash in this mica molecule 

 the excess of alumina must be drawn from an excess of alumina 

 over the alkalis originally in the rock, or it must be drawn 

 from a sodium-aluminium molecule, since no potash molecule, 

 in which potash exceeds alumina, occurs to any extent in the 

 pyrogenetic minerals according to present theories. In the 

 second instance the effect would be to reduce the albite or 

 nephelite molecules and drive the soda into combination with 

 ferric iron, affecting the distribution of the silica. 



The development of the metasilicate molecule, Na Al Si 3 6 , 

 which enters amphibole, would reduce the albite molecules with 

 the liberation of silica and the possible production of quartz, or 

 the raising of nephelite to albite. It would reduce the nephelite 

 molecules if it did not affect those of albite, and would require 

 additional silica, which would have to be drawn from a polysili- 

 cate (orthoclase) or from a metasilicate. It does not appear to 

 enter largely into the composition of rock-making amphiboles, 



