246 STRATIGRAPHY 



granite of Depot Island. They comprise typical gabbros with sphene, sphene-bear- 

 ing amphibolite, hypersthene gabbros, bronzite gabbros, and uralite porphyry. The 

 following photograph shows the relation of the inclusions of basic sphene-bearing 

 diorites at Depot Island : — 



2a. Grey granites with their granophyres, granite por2)hyries, quartz porphyries, 

 aplitic granite porphyries, pegmatites and aplites. The quartz and felspar por- 

 phyries, aplites, and pegmatites, are obviously a little newer than the granites. (See 

 Fig. 37.) 



2&. Pink Granites. No actual junction was seen by us between the pink 

 granite and the grey granites, but nevertheless, for reasons already given, we agree 

 with Ferrar in considering the pink granite to be newer than the grey. 



3. Dyke rocks of basic type, such as lanvprophyres, minette, kersantite, d:c. These 

 basic dykes appear to belong to a newer cycle altogether than that which has 

 produced the older diorites, gabbros, and grey and pink granites. The well-marked 

 quartz and felspar porphyry dyke at Backstair Passage, near Mount Larsen, may 

 perhaps be a complementary dyke rock in the kersantite group. A feature of 

 special interest in the diorite and gabbro group is the abundance of sphene. 



The analysis shows that TiOg is present to the amount of 2*34 per cent. Rolled 

 crystals of titaniferous iron were observed by us in erratics of quartzite at Cape 

 Royds. These evidently were derived from some old rocks rich in titanium, possibly 

 the diorites or gabbros already mentioned. 



It is also interesting to note that even at this early epoch the plutonic rocks 

 show a tendency to being alkaline. The grey granite, for example, of Mount Larsen 

 contains l^a,.fi 3-13 per cent., Kfi 2-87 per cent., while the pink granite of Cape 

 Irizar contains NajO 3 "8 8 per cent., KoO 472 per cent. 



There is thus a tendency for the newer granite to be somewhat more alkaline 

 than the older. This region was thus early marked out as a province for the 

 alkaline rocks. 



The interesting question now suggests itself, do these plutonic and hypabyssal 

 rocks show any distinct affinities with the rocks of the South American Andes? 

 In his account of the petrology of the specimens collected by Dr. Charcot's Exjjedi- 

 tion, M. Gourdon emphasizes the fact, to which Dr. Otto Nordenskjtild has already 

 called special attention, that on the American side of the Antarctic the petrological 

 characteristics of the Andean chain of South America are distinctly repeated. The 

 rocks from that locality show predominance of lime with a little soda, so that the 

 felspars are essentially lime-soda felspars rather than soda-potash. The felspars 

 also exhibit very well the zoning which is so characteristic of Andean rocks. 

 Gourdon records the occurrence of one rock only — an erratic of micro-granite with 

 soda pyroxene and soda hornblende — which is alkaline in character. In regard to 

 this he makes the following statement : — 



" Ne serait-elle pas I'indice de I'existence plus au sud d'une province petrographique 



