III.] ORIGIN OF SANTA CRUZ FAUNA. 125 



This, however, is a difficult and perplexing subject which it is 

 scarcely possible to explain fully in the present imperfect state of 

 palaeontological knowledge. 



With regard to the marsupials of an Australian type, it has 

 already been stated in the preceding chapter 1 that these appear to 

 have been derived from Notogaea by means of a southern land- 

 bridge. The hypothesis there suggested is that the immigration 

 has taken place via the Antarctic continent, probably across the 

 southern Pacific 2 . It is known that shallow water extends from 

 southern Patagonia andTierra del Fuego to South Shetland; while 

 between Australia and the Antarctic land there are no depths ex- 

 ceeding 2000 fathoms. On the other hand it is just possible that 

 the connection may have been by way of Polynesia. 



In this place reference may be made to certain very remark- 

 able resemblances existing between animals of groups other than 

 mammals respectively inhabiting Neogsea and Notogsea. The first 

 instance is that of two peculiar families of freshwater fishes, known 

 as the Haplochitonida and Galaxiidce. Of these the former has 

 one Australian and a second South American genus, while the 

 latter is represented by a single genus (Galaxias), common to 

 New Zealand, Australia, and the extremity of South America. 

 This, however, is by no means all, since one species of the last- 

 mentioned genus (G. attenuatus) is found in regions as remote from 

 one another as New Zealand and Tasmania on the one hand, and the 

 Falkland Islands and the extremity of Patagonia on the other. 

 Commenting on this, Dr Wallace remarks 3 , it is impossible to 

 believe that a land connection between South America and 

 Notogaea could have existed " within the period of existence of 

 this one species of fish, not only on account of what we know of 

 the permanency of continents and deep oceans ; but because such 

 a connection must have led to much more numerous and im- 

 portant cases of similarity of natural productions than we actually 

 find. Rather must we look to the transport of the ova across the 



1 Supra, p. 55. 



2 Possibly, as suggested below, the connection may have been nearer the 

 tropics. 



3 Geographical Distribution of Animals, Vol. I. pp. 401, 402; see also Vol. 

 II. pp. 82, 83. 



