/'T'i H. S. Sumiiiers: 



1. Dacite, Cherokees, Macedon District. 

 II. Dacite, Willimig-ongong Creek, Macedon District. 



III. Dacite, Braemar House, Macedon District. 



IV. ' Dacite, Ujjwey, Dandenong- District. 



V. Granodiorite, Braemar House, Macedon District. 

 Vr. Dacite, Hesket, Macedon District. 

 VII. .\dainellite, Broadmeadows. 

 VIII. Granodiorite, Hesket, Macedon District. 

 IX. idamellite, Trawool, Strathbogie District. 

 X. Average of Adamellite and Quartz-porphyry, Strathbogie 



District. 

 XI. Granodiorite, Harcourt. 

 XII. Quartz porphyrite, Violet Town, Strathbogie District. 



XIII. Adamellite, Ingliston. 



XIV. Granodiorite porphyry, Barringo Creek, Macedon District. 

 XV. Granite, Gabo Island, 



XVI. Granite, Mount Buffalo. 

 XVII. Glranite, Dog Rocks, Geelong. 

 XVIII. Granite, Cape Woolamai. 



If the variation diagram, Fig. 3, can be couhidered at all a 

 picture of the nature of the differentiation of the magma which pro- 

 duced the various rock types shown, then it is certain that there 

 must be limiting values beyond which differentiation cannot pro- 

 ceed. It is obvious that for any percentage of silica that percent- 

 age plus the percentages of the bases as read off on the diagram 

 should in all cases equal one hundred. It follows then that when any 

 line reaches the vanishing point, i.e., the percentage of the oxide 

 represented by that line becomes nil, differentiation as indicated by 

 the diagram must cease. The reason for this statement is that 

 beyond that point the silica percentages plus the percentages of the 

 bases, as indicated on the diagram, must be more than one hundred 

 by the amount (a minus quantity) read off from the vanished line. 

 In order to bring out this point clearly a fresh diagram has been 

 drawn (see Fig. 4). 



Two extreme values have been taken tu plot the iliagram, viz.. 

 the mean of the three dacites from Cherokees, Willimigongong Creek 

 and Braemar House, and the mean of the granites from Dog Rocks 

 and Cape Woolamai. The lines have been drawn exactly through 

 these points so that the composition of any possible rock as read off 

 from the diagram may equal one hundred per cent. 



The two values used are as under. The molecular ratios are given 

 in the second column in each case. 



