228 JOSEPH P. IDDINGS 



the quartz limit still further toward the right, in the direction of 

 lower silica percentages. Quartz must therefore occur in rocks 

 with lower and lower silica in proportion as the alkali-silica 

 ratio decreases, and as lower silicate molecules are developed 

 instead of higher silicates. The line AC is probably very near 

 the highest limit of no quartz to be found in noncrystalline 

 igenous rocks. 



If any of the iron crystallizes as magnetite instead of com- 

 bining with silica to form ortho- or metasilicates, silica will be 

 liberated to form more albite out of nephelite, or to produce 

 more quartz by the side of albite. The effect of the develop- 

 ment of magnetite then is to shift the quartz-nephelite limit 

 farther to the right, that is, toward lower silica percentages. In 

 the case of rocks within the nephelite limit, the effect will be to 

 decrease the nephelite and increase the albite. It may also 

 react upon the ferromagnesian silicates when orthosilicates are 

 present, permitting the liberated silica to convert some of the 

 orthosilicates into metasilicates without materially affecting the 

 relative proportions of the other constituent minerals. 



In this connection attention may be called to the line NFa, 

 which is the range of magmas composed wholly of nephelite 

 and the orthosilicate of iron. It appears to be the limit toward 

 which the highly sodic magmas tend, as may be seen in 

 Diagram I. Magmas low in alkalis associated with the erup- 

 tive iron ores lie beyond this boundary. 



II. In order to discover the relations that would obtain for 

 magmas in which the alkali is wholly potash when alumina is present 

 iti amounts just equal to the potash, we have to consider the range 

 of positions for rocks corresponding to mixtures of quartz and 

 orthoclase (potassium-aluminium polysilicate), and of orthoclase 

 and leucite (potassium-aluminium metasilicate). This corre- 

 sponds to the red line QOL. As already remarked in the 

 previous paper, the lowest potassium-aluminium silicate among 

 the feldspathoid minerals appears to be the metasilicate, leucite, 

 and when potassium is found in rock-making nephelite it may 

 be explained as in a metasilicate molecule mixed with the 



