112 ME. C. E. TILLET OX THE GRAlSriTE-GjS'EISSES [vol. IxXTli, 



triangular diagram. It will suffice to take, in addition to the 

 amphibolite analysis, those of the quartz-gabbro of Georgia and 

 the diabase of Somerset County (New Jersey). 



The parameters of these rocks, when calculated, are : — 



a c f 



(i) Point Boston ampliibolite 3"6 4*9 11"5 



(ii) Georgia quartz-gabbro 3"9 5*2 10*9 



(iii) New Jersey diabase 3"7 4*5 11*8 



In the accompanying diagram, the positions of these rocks are 

 indicated b}' the points 1, 2, and 8, respectiyely (fig. 8). In plain 



Fig. 8. 



-Osann -projection of tlie Boston ampliibolite,' and 

 igneous rocks conqjarahle thereivith. 



circles are placed the mean positions of Groups III and lY of 

 Grubenmann's classification. From this it is seen that the analyses 

 group themselves in proximity to the mean position of the plagio- 

 clase-gneisses (HI), and thus show some affinity'' to the dioritic 

 group of igneous rocks. 



Attention may here again be drawn to the description of an 

 amphibolite-band in the Kirton-Point area, which is a foliated 

 amphibolite at the border, but near the centre shows a much more 

 massive structure, and is in fact a partly-uralitized micropegmatitic 

 dolerite (p. 102). There is thus both field and chemical evidence to 



