Geoloyn of Li/ifdalr. 365 



plutdiiic. and hypalAssal phases of the eyele, and they occurred in 

 that ordiT. 'J'hr (.hriuical composition of these three types is 

 practically the same. It is clcai'. then, that after the extrusion of 

 the tirst four- series of the volcanic cycle, the magma remained 

 unchanged, not only throughout the last series of the volcanic phase, 

 though that series probably exceeds in magnitude all the previous 

 ones added together. l)ut also throughout the plutonic and hypa- 

 byssal phases to the cud of the cycle. 



2. These three types also l)ear the same relationship to one 

 another in the Maccdon district, wliere no differentiated rocks of 

 tlie dacite family have l)een found. 



.'1 There are immense developments of normal dacite, sometimes 

 associated with granodiorite, in the neighbouring mountains be- 

 tween Healesville and Warburton, the Black's Spur, and the Cer- 

 berean Ranges. 



With these reasons for regarding the Upper Dacite as representing 

 the undifferentiated magma, there remain but four volcanic series 

 whicli are differentiated, viz., a, b, c, and d. 



By both mineralogical and chemical evidence these four series are 

 resolved into two pairs. In the first pair (a and h) quartz is absent, 

 except in the ground mass. In tlie second pair (r and d) it is abun- 

 dant. In the variation diagram, the second pair fall with the 

 Upper Dacite (e), the Granodiorite, and the Wandin Dyke, on the 

 same curves. But the first pair fall on separate curves, being richer 

 in SiU^,. AI0O3, and alkalis, but poorer in oxides of Ca, Mg, and 

 Fe, than the second pair. 



It thus appears that, if the toscanite be admitted to this cycle, 

 there must have been an early separation of the magma into two- 

 parts. 



One of these redivided, forming two magmas (a and b), almost 

 identical in chem. composition, except that a was much richer in 

 HrjO than 6, with the result that though chemically similar, they 

 are mineralogically different. 



The other part, which was richer in basic constituents, and 

 though poorer in SiO^, contained much free quartz, divided into 

 two series (c and d), such that the latter, containing more of the 

 basic constituents, and more water, produced a rock more crystal- 

 line than the former, and containing less quartz, but more felspar 

 and ferro-magnesin minerals. 



Examples of the differentiation of andesitic magmas have been 

 described by Judd (21), and by Teall and Lagorio. The latter 

 observers remark that the older of crystallisation is such that the 



