312 J. A. THOMSON. 



tion between the group and the Mesozoic dolerites of 

 Tasmania before alluded to. Further, the dolerites and 

 quartz dolerites of Victoria Land, Antarctica, also possess, 

 according to the writer's observations, the same peno- 

 logical peculiarities, and in particular the minerals enstatite- 

 augite and pyroxene-perthite. Though the writer's obser- 

 vations on the Karroo dolerites of South Africa have not 

 been sufficiently extensive to allow the same affirmation to 

 be made, Wahl has shown 1 that the ' diabase ' of Richmond 

 originally described by Cohen 2 contains enstatite-augite, 

 and it is reasonable to suppose that this mineral has a wide 

 occurrence in South Africa. 



The Tasmanian and South African rocks are definitely 

 known to be of late Mesozoic or early Tertiary age. There 

 is no evidence so far produced to show that the others are 

 not of the same age. They all occur in the remaining 

 horsts of the foundered Gondwana Land, and the question 

 arises whether they do not point to the presence of an 

 immense magma or to a series of similar magmas, which 

 in Mesozoic times underlay the old Gondwana Land, parts 

 of which were forced up by the earth movements which led 

 to the breaking up of the continent. 



Mr. Benson has kindly drawn my attention to a paper in 

 which Prior 3 has already made a similar suggestion, i.e. he 

 suggests that the dolerites of Zululand and Victoria Land 

 are of the same age, and further points out the curious 

 association of these rocks on the mainland in each place 

 with later alkaline rocks in outlying islands. This obser- 

 vation might, with reservation, on the island-occurrence 

 of the alkaline rocks be equally applied to Tasmania. But 



1 Loc. cit., pp. 29-30. 



8 Geognostisch-petrographische Skizzen aus Sud- Africa, Neues Jahr. 

 f. Min. etc. 1887, B.B., p. 234. 



3 Prior, G. T., Petrographical Notes on the Dolerites and Rhyolites of 

 Zululand, Ann. Natal Mus.n, 1910, p. 152. 



