Repay t on the Rock-Specimens. 331 



Slates and quartz-grits appear to be the prevailing rocks beneath 

 the basaltic lavas over a large part of Victoria Land. They consti- 

 tute the main-mass of Duke of York Island, and occur on the 

 opposite mainland (Geikie Land). According to Mr. Bernacchi, the 

 junction of the slates and the basalts occurs at a place about two miles 

 south of Duke of York Island in the bottom of Robertson Bay, where 

 the basalt is seen to liave flowed over the sedimentary formation and 

 hidden its southern prolongation. To the east of the junction extends 

 the basalts forming the mass of Cape Adare, while along the coast to 

 the west towards Cape North occur only slates. That the slates 

 continue southwaid underneath the basalts towards Mount Erebus 

 is considered probable by Mr. Bernacchi, since slates were seen Ijy 

 him at the foot of Mount Melbourne in Newnes Land. 



As regards the basalts, apart from the dyke rocks, the lavas of 

 Cape Adare, Possession Island, Coulman Island, and Mount Terror 

 are all very similar fine-grained glassy l)asalts, characterised by the 

 presence of basaltic hornblende, so that it is reasonable to suppose 

 that, as Mount Erebus still shows signs of volcanic activity, they 

 have all been erupted at a comparatively recent date. 



As to the connection between the phonolitic rocks and the 

 basalts, their presence together in the tuff described on p. 330 

 suggests that they belong to one period of eruption ; and it is pro- 

 bable, from the result of the analyses given on ])p. 326 and 329, that 

 both are differentation products of one magma, fairly rich in soda and 

 poor in magnesia. Such a magma might have the composition of a 

 basic nepheline-syenite, like that which helps to form the core of 

 Mount Kenya (see Gregory, 'Quart. Journ. (ieol. Soc.,' Ivi. (1900), 

 p. 208), or it might possibly be represented by the olivine biotite 

 rock described on p. 323. 



An important point to be considered in the examination of this 

 collection is the light it may throw upon the supposed connection 

 betAvcen the Antarctic lands and the neighljoiiring continent of 

 Australia. Prof. J. W. Gregory, in 'Nature' (Vol. Ixiii., p. 609), 

 refers to Kitter's suggestion that Wilkes Land is a southern extension 

 of the Australian plateau, and that the volcanic chain of Victoria Land 

 is the continuation of the volcanic line which passes through New 

 Guinea, New Caledonia, and New Zealand. As regards the more 

 ancient rocks in the present collection, in support of this suggestion 

 is the fact that they are very similar to, and are even stated by 

 Prof. Gregory to be " practically identical with some of the Lower 

 Palaeozoic rocks of Victoria" ; while in the case of the more recent 



