Vol. 50.] BASIC ERUPTIVE BOCKS OF GRAN. 31 



that tlie above-mentioned kinds of rock are otherwise connected 

 with nepheline-syenites. With the '■hern' hypothesis of Rosenbusch, 

 on the other hand, the above proved connexion does not seem to agree. 



[Note. — For rock-types, differentiated out of a common magma, I 

 propose the name 'complementary rocks'; camptonites and bostonites 

 are, then, such complementary rocks. Between the dyke-rocks we 

 have also a number of other examples. 



Complementary rocks should, therefore, in the classification of 

 rocks be placed in that rock-group which has a chemical composition 

 agreeing with the composition of the original common magma of the 

 complementary rocks ; minettes and aplites, for instance, are com- 

 plementary rocks of the granite-family, and so on. — Dec. 30th, 1893.] 



VI. Differentiation in the Bosses. 

 The differentiation of the olivine-gabbro-diabase magma into 

 camptonites and bostonites is not the only one which has taken 

 place in this magma. In the boss of Brand berget we find that the 

 same or a very closely allied magma has, by differentiation uuder 

 other conditions, given rise here to other products. It is true 

 that a magma of hornblcndic composition (partly crystallized as pure 

 hornblendite, partly as a camptouitic rock) was differentiated out 

 from the original magma ; but these masses are here quite unim- 

 portant. The conditions here have evidently not permitted to any 

 great extent the crystallization of dark alumiuiferous minerals, such 

 as brown hornblende ; consequently the differentiation of the magma 

 has not allowed the concentration of liquid compounds of analogous 

 composition along the margins of the boss. On the other hand, it 

 seems clear that the ruling conditions have highly favoured the 

 crystallization of dark pyroxene, rich in calcium, magnesium, and 

 iron, and relatively poor in aluminium. Agreeably to this supposition 

 we find that, along the margins of the boss, especially in the west 

 and north, there has been differentiated a basic magma of an 

 almost pure pyroxenic composition, which has often crystallized as 

 very coarse-grained pyroceenite, with as much as 95 per cent, 

 pyroxene. An analysis from the laboratory of Herr L. Schmelck 

 gave the following composition for such a very coarse-grained 

 pyroxenite from Brandberget : — ■ 



Table VIII. 



SiO, .... 



Tio;.... 



Al.,<5 8 • 



F e ;o 3 . 



FeO .... 

 MgO.... 

 CaO .... 

 Na 0.... 

 K 2 .... 

 Ca 8 P 2 8 

 E.,0 .... 



Pyroxenite, 



Pyroxene, 



Brandberget. 



Limburg. 



45-05 



44-65 



2-65 



2-93 



650 



662 



3-83 



5 02 



7-69 



3-87 



1207 



14-76 



18-66 



2032 



094 



1-29 



0-78 



049 



0-31 





240 





100 88 99-95 



