Victorian Igneous Rocks. 277 



Turing now to the more basic liypothetical limit of possibla 

 differentiation, the composition of the rock may l)e estimated in a 

 similar manner, an equal amount of fci-rous iron l)ein«r allowed 

 for magnetite as is re<|uired foi- ilniuiiite. 



This composition seems a rather more rational end point than the 

 other, as the rock would consist essentially of two minerals, anor- 

 thite and aluminous amphibole or pyroxene. 



It is interesting to note that the ratio of the quartz to felspar is 

 39.3; 60.7 in the case of the acid limiting value, and that the 

 ratio of the pyroxene to felspar in the case of the basic limit is 

 approximately the same, viz., 41.1; 58.9. 



While suggesting that these two values as read off from the 

 diagram represent the theoretical limits of this linear differentia- 

 tion, it does not follow that rocks w-ith such composition would 

 necessarily separate out from the given magma. In actual occur- 

 rence the composition of the original undifferentiated parent 

 magma must be a controlling factor in limiting differentiation. 



(c) Macedon Kainozoic Alkali Series. 



All the leading types of rocks belonging to this series have been 

 analysed at the Mines Department Laboratory, and in all nineteen 

 first-class analyses have been made. These are given in Table XI. 



* This is not a true nietasilicate as the AI2O3 and SiOj are combined in the proportion 1 : 1, but 

 an examination of a number of analyses of aluminous ampliiboles and pyroxenes shows that gen- 

 erally silica must be allotted to alumina in the above proportion. 



