Summary of Results.] SUB ANT ARCTIC ISLANDS OF NEW ZEALAND. 805 



runs northward towards Gough Island and Tristan da Cunha, and another arm 

 extends east past Bouvet Ishxnd and the South of Africa towards Madagascar. The 

 Prince Edward and Marion Islands and the Crozets would apparently be also con- 

 nected with this ridge. 



It is important to note that these soundings seem to show how connections 

 could be made between the Antarctic Continent and South America and the sub- 

 antarctic islands to the south of the Indian Ocean by local elevations of the land of 

 comparatively small extent and without assuming any very great change in the 

 relative positions of the great continents and oceans. These would supply the 

 connections required by Forbes without assuming such a huge subantarctic con- 

 tinent as appeared to be necessary at the time when he wrote. 



At the present time the Antarctic Continent proper is almost devoid of terrestrial 

 life. The plants are represented by one species of flowering-plants {Deschamfsia 

 antarctica), and by a few minute mosses and lichens ; while the land-animals are 

 confined to a small number of wingless insects and spiders. Evidently the Antarctic 

 Continent theory requires a warmer climate for antarctic land in past ages, and evi- 

 dence for this is now available. The National Antarctic Expedition found fossil 

 leaves near the winter quarters of the " Discovery," though these were unfortunately 

 not sufficiently well preserved for exact identification ; seams of coal were afterwards 

 found by Lieutenant Shackleton still further south. The most important discovery, 

 however, was that made by the Swedish Antarctic Expedition, of abundant fossil 

 plants in a sandstone formation of Tertiary age on Seymour Island. These have 

 since been investigated by P. Dusen, and, though his report is not available in New 

 Zealand at the present time, abstracts of it which have appeared are sufficient 

 to show the very important bearings that this discovery has on the question now 

 under discussion. Among the fossils are specimens referred to a large number of 

 genera, such as Laurelia, Drimys, Ilicifhyllum, Lomatia, Knightia, Fagus, Nothofagus, 

 Araucaria, &c. This Tertiary flora of Seymour Island appears to be a mixture of 

 temperate and subtropical species, and is distinctly related to those of two present 

 South American floral regions — namely, the temperate flora of south Chili, and 

 the subtropical plant - world of south Brazil. Its connection with Australia and 

 New Zealand is evidently greater than Dusen appears to think it. The presence 

 of Knightia Andreas, a close relative to K. excelsa of New Zealand, and of such 

 genera as Nothofagus, Drimys, and Laurelia shows very considerable connection 

 with the present plant covering of New Zealand. 



In this connection it should be mentioned that Ortmann (1904) has demon- 

 strated the existence of fossils in the Lower Miocene beds of Patagonia, which contain 

 Mollusca allied to those found in Chili, New Zealand, and Australia, and has pointed 

 out that these species of Mollusca belong to a littoral fauna. Unfortunately, Ort- 

 mann's paper is not here available, and I do not know what species are represented 

 in these Patagonian beds ; but the existence of the same Tertiary fossils in beds 

 in South America and New Zealand was already to some extent known, and Mr. 

 Suter has given me a list of six species common to New Zealand and South America.* 

 The existence of this littoral fauna in Lower Miocene times implies the existence 



* Epitonium rugulosum lyratum, Zittel ; Crepidula grcgaria, Sow. ; Turritella ambulacrum, Sow. 

 CucuUaea alta, Sow. ; Venericardia patagonica. Sow. ; Brachydontes magellanica, Lam. (= Mytilus). 



