32 PROF. W. C. BKOGGEE ON THE [Feb. 1 894, 



For purposes of comparison the analysis by Merian (Neues Jahrb. 

 1885, Beilage, vol. iii. p. 285) of the pyroxene from Limburg, Kaiser- 

 stuhl, Baden, is placed beside that of the Brandberget rock. The 

 agreement between the rock and the mineral is, as will be seen, very 

 close. 



The typical pyroxenite of Brandberget consists, practically in its 

 entirety, of dark pyroxene ; mingled with this, but quite subordi- 

 nate, are a little brown hornblende, reddish-brown biotite, traces of 

 plagioclase, etc. ; the structure is often miarolitic, with various 

 minerals (splendid crystals of titanite, apatite, etc.) in the open 

 cavities. 



This pyroxenite is traversed by innumerable, comparatively acid 

 veins of fine-grained, light-grey augite-diorite, or mica-augite- 

 diorite, invariably rich in yellow titanite, and of a kind passing into 

 the series of augite-syenite, designated formerly by the author as 

 ' akerites.' These veins of augite-diorite are so abundant that 

 the rock on the whole of the western side of Brandberget represents a 

 typical eruptive breccia, containing angular fragments of the dark 

 coarse-grained pyroxenite cemented by the fine-grained, light-grey 

 augite-diorite. This latter then represents the acid residuary 

 magma after the differentiation of the pyroxenite-magma, already 

 consolidated to pyroxenite, when its veins were squeezed up. 



The main mass of the magma of Brandberget is not much 

 differentiated, but has crystallized out like the basic olivine-gabbro- 

 diabase mentioned above, differing only slightly from the above 

 calculated average magma of olivine-gabbro-diabase. 



All the products of crystallization on Brandberget represent, 

 then, the results of a special differentiation, which has taken place 

 in this boss itself. As we found above (p. 27) that 9 parts of the 

 camptonite composition, and 2 parts of the bostonite composition, 

 gave us the average composition of the olivine-gabbro-diabases of 

 Gran and Modum, in a similar manner it should be possible by the 

 mixture of some parts of pyroxenite, hornblendite, augite-diorite, 

 and olivine-gabbro-diabase (of the analysed composition) to re- 

 construct an average magma very closely allied to the average 

 composition of the olivine-gabbro-diabases of Gran, or perhaps 

 slightly more basic. To establish this calculation I still need 

 several analyses; a preliminary trial has shown the approximate 

 proportions, and these proportions seem to agree with the obser- 

 vations made in the field as to the extension of the different kinds 

 of rock on tho hill-top. When, in a monograph on these basic rocks, 

 I can present, as I hope to do, the exact calculation, founded on 

 a sufficient number of analyses, showing that cc parts of the analysed 

 basic olivine-gabbro-diabase of Brandberget mixed with y parts 

 of hornblendite, z parts of pyroxenite, and w parts of augite-diorite 

 (all of analysed varieties from Brandberget) give about the average 

 composition of the olivine-gabbro-diabase above calculated from the 

 analyses of rocks from Brandberget, Solvsberget, and Dignaes, then I 

 hope to have proved that all the various kinds of rock from Brand- 

 berget must be considered as products of differentiation from a 



