294 H. S. Suitimers : 



The evidence of the Geehjiig and Heathcote areas suggest that 

 magmas of intermediate coiiipositiou have split into two portions, 

 an acid and a basic. We have also seen that a basaltic magma 

 cannot be serially related to the Victorian granite-dacite series. 

 This suggests the hypothesis which is tentatively advanced that 

 wherever a molten magma of intermedite composition exists, 

 Avhether it be a primary magma or a secondary magma derived 

 by the fusion of pre-existing igneous and sedimentary rocks, then 

 the primary differentiation is into an acid portion and a basic 

 portion, possibly due to the limited miscibility of the two types. 



Basaltic magmas appear to have a greater uniformity in com- 

 position, and to produce fewer subsidiary types, to which they are 

 serially related, than the magmas which have produced granite, 

 dacite, etc., so that possibly the governing factor in this primary 

 differentiation is the separation out of a basic magma of fairly 

 constant composition. 



If this primary differentiation were incomplete, or if by assimi- 

 lation of other material the balance between the various constituents 

 in the basaltic magma was upset, then secondary differentiation 

 might take place, causing a separation out of a subsidiary magma 

 and leaving a normal basaltic magma. 



If the subsidiary magma were molecularly unstable, then further 

 differentiation of the serial or complementary type would be 

 induced, giving rise to a varied facies such as is met w4th among 

 the Macedon alkaline rocks. 



In the case of the acid primary differentiation product, judging 

 by the Victorian examples quoted the various types would be 

 evolved by serial differentiation, with possibly some minor com- 

 plementary differentiation as at Lilydale. 



As in the case of the basic primary differentiation product, if 

 the molecular balance be upset either by incomplete primary 

 differentiation or by assimilation, then a subsidiary magma may 

 separate out and give rise to rocks less normal in character. 

 Basic segregations so conmion in the Victorian granites may be 

 regarded as possibly representing such subsidiary differentiation 

 caused by the magma assimilating foreign material during intru- 

 sion. 



As the acid primary differentiation product would have a lower 

 specific gravity than the basic portion, we might expect the acid 

 portion to occupy the upper portion of the primary magma reser- 

 voir, and in case of crustal movements, due to lateral compiession, 

 might bo more affected than the lowei- basic portion. On the other 



