88 ALFRED C. LANE 



about this in their slag formulas. Consequently, it is a priori 

 probable that labradorite and augite magmas mix readily and 

 without tendency to split; though, when they actually come to 

 crystallize, petrographic observation teaches that the labradorite 

 feldspar is the earlier and the augite later for the rocks on the 

 right or lower silica side of the average rock. The reverse is 

 true on the other side. We must remember that, so far as 

 chemical affinities allow, the eutectic (most fusible, least solidi- 

 fied) magma will contain a little of every element going; and 

 therefore we are not surprised at the complexity of the composi- 

 tion of the augites which are collected in the table accompany- 

 ing the statement of the Quantitative System. The latest researches 

 show that basaltic magmas are fusible at distinctly lower tempera- 

 tures than either labradorite or augite. 



Nor are we surprised that, if we draw a line from the average 

 rock-analysis point, or that for labradorite, in the average direc- 

 tion in which the less silicious and less alkaline analyses diverge 

 therefrom, it will strike the line of no alkalis at 43 per cent. 

 SiOg, or somewhat less; for this is not only the lowest silica 

 percentage which augite, diallage, and the feldspar series reaches 

 (anorthite), but the maximum for olivine. 



So in the femic magmas there is a clear tendency away from 

 extremely different percentages of iron, lime, or magnesia. 

 Probably the ratio CaO : MgO : FeO : : 2 : i : i is not far from 

 the eutectic one. In the following group of analyses from 

 Lighthouse Point, for instance, it is clear that at the center, 

 where there was opportunity for differentiation and adjustment 

 of the magma, and the eutectic would accumulate, there is more 

 lime than at the quickly chilled and cooled contact. The magma 

 had an excess of magnesia and iron for eutectic relations, and 

 this is shown petrographically by an early generation of olivine 

 and magnetite. 



Returning to the Iddings's diagrams, it seems, from a study of 

 the distribution of Class I, that a few rocks are included 

 (anorthosites, canadases) which it goes against the grain to 

 include — rocks with an unusual amount of lime, but really tribu- 

 tary to the femic eutectic ; that the real natural family is from 



