234 JOSEPH P. IDDINGS 



Let us consider cases I and II. The molecule (Mg,Fe)(Al, 

 Fe) 2 Si0 6 is a subsilicate. The range of silica percentages for 

 all manner of variations in its composition is from 29.52 for 

 MgAl 2 Si0 6 to 20.54 for FeFe 2 Si0 6 . The molecule usually is 

 more nearly the average of these extremes. The silica range is 

 indicated in Diagram I by the bracket, which shows that this 

 molecule requires less silica than the orthosilicate of iron, con- 

 siderably less than olivine, and very much less than diopside or 

 enstatite. Its development, then, would extend the quartz limit 

 and reduce the nephelite-leucite limit very considerably in pro- 

 portion to its own amount. It may be stated that in general its 

 amount is quite subordinate to that of the ferromagnesian ortho- 

 and metasilicates. Its effect on magmas of case III, where 

 anorthite occurs, would be of the same kind as the introduction of 

 non-aluminous ferromagnesian silicates, but the degree of the 

 change would be greater in proportion to the number of mole- 

 cules introduced. 



Turning to the alkali-ferroaluminous molecules, it is to be 

 observed that one, (K,H) (Al, Fe) Si 4 , is well known to be 

 capable of appearing in igneous rocks of almost any chemical 

 composition in which its elements occur. Two others (K,H) 

 AlSi0 4 , and NaFeSi 3 O e appear to occur in rocks with special 

 chemical composition, while the fourth, NaAlSi 2 O e , is less well 

 known but may be supposed to be of wide occurrence in small 

 amounts. The first belongs to biotite. If the ratio between 

 potassium and hydrogen varies from 1 : I to 1 : 2, and the alumin- 

 ium and iron be supposed to reach their possible extremes, 

 from normal biotite to lepidomelane, the range in silica percent- 

 age for this orthosilicate molecule would be from 29.07 to 19.10, 

 which is indicated by the bracket in Diagram 1. The relative 

 proportions of potash and alumina vary from I : I to 1:2 in 

 most instances. Sodium does not enter into the molecule to 

 any considerable extent, the development of biotite then affects 

 the potash-feldspathic molecules. In orthoclase rocks within 

 the quartz limit its development would reduce the amount of 

 orthoclase molecules, and would change a polysilicate into an 



