40 



Again, it has been shown that after the granitic stage of 

 a magma is reached differentiation may be prolonged so as to 

 produce soda granites and syenites through the impoverish- 

 ment of the magma in potash and silica, (^2) and possibly 

 through the concomitant concentration of mineralizers. Here, 

 as there is an actual decrease in silica with advancing differ- 

 entiation, the sequence could not be represented in its entirety 

 on the ordinary variation diagram, and it really seems as if 

 there is no convenient method, under these circumstances, of 

 giving graphical expression to the complete differentiation- 

 history of the magma. 



(4) SHAPE AND NATURE OF THE INTRUSION. 



Only very scanty data are obtainable as regards the 

 formal relationships of the rock mass. It is to be remem- 

 bered that the few outcrops available at the present day 

 probably represent, only a small marginal remnant of a once 

 very extensive intrusion, and that the only place where the 

 actual junction with the country rock may be observed is at 

 Rosetta Head. It is therefore evident that there can be 

 little certainty as to the original shape and the exact nature 

 of the mass. 



The boundary between granite and schist is interesting. 

 From the photograph of Rosetta Head it will be seen 

 that the junction, which has been sharply defined through 

 differential erosion, makes an angle with the horizontal vary- 

 ing from about 13° to 45°, and that the bounding surface 

 dips in towards the granite instead of off it, as is the usual 

 case. The invaded schists dip. about S. 10° W. at angles 

 varying from 35° up to 60°. The local variations in 

 direction and amount of dip are doubtless an effect of the 

 intrusion. The evidence points to the series of rocks having 

 been formed almost entirely by differentiation of a deep- 

 seated or intratelluric magma and its subsequent eruption, in a 

 partly crystallized condition, to a higher level in the litho- 

 sphere. The different rock types show genetic relationships 

 towards each other; successive injections of magma at short 

 intervals are indicated by "swirling" contacts, as between 

 diorite and porphyritic granite, or between porphyritic and 

 even-grained granites; while the presence of "rifted blocks" 

 of schist gives evidence of the force with which the magma 

 thrust itself up through the older rocks. The intrusion then, 

 whatever its shape, is a multiple one, and the whole 

 assemblage of rocks may be regarded as a plutonic complex 

 in Harker's sense, the component members of which have 



(32) Vide Bowen : Jour. Geo!., vol. 33, 1915, Supplement, 

 p. 5-3, et seq. 



