,328 FOUMKR LAND CONNECTIONS. 



_ extremity of South America, J^'alkland Islands, Southern 

 Australia, Tasmania and New Zealand. They are comparatively 

 primitive fishes according to Mr. Regan, and to that extent 

 lend credence to the theory of Nichols and Griscom, that primi- 

 tive freshwater fishes origmated in the south, and the more 

 recent fishes in the north. Some of the species resort to the 

 sea for breeding purposes, and one species was originally 

 described as marine, but this now seems to be incorrect. But, 

 on such grounds apparently, they were regarded both by 

 Boulenger and Regan as of marine origin. The South African 

 species are quite distinct from any other, and there is no evidence 

 that any species can cross wide oceans, although one species is 

 common to South America and Australia. No fossils are known. 



(2) Hie distribution of the freshwater Isopod Crustacean known 

 as PJireatoicus- is most suggestive. It occurs in the Western Cape 

 Province (as made known by Mr. K. H. Barnard), in Australia, 

 Tasmania, and New Zealand;' in fossil form it has been described 

 by Prof. Chilton'^ from Upper Triassic rocks in Australia. If 

 tliis genus has been cosmopolitan during its prolonged life, it is 

 a strange coincidence that the only place where it is known in 

 fossil form is the region where it occurs to-day. 



(3) The distribution of penguins is comparable to that of 

 the genus Galaxias and the common species of the Cape has its 

 nearest relative in the Falkland Islands. In the northern hemi- 

 sphere penguins are quite unknown, either living or fossil, but 

 they are known in fossil form from New Zealand and Patagonia, 

 and are said to occur in early Tertiary rocks on Seymour Island. 

 Thus we see that they have lived in the southern hemisphere for 

 a very long period ; and nothing can be stated in favour of their 

 northern origin. This point needs emphasis because most writers 

 who object to hypothetical land-bridges are obsessed with the 

 idea that all the main groups of aniulals actually originated 

 in the northern hemisphere. 



(4) The distribution of the Cystignathidac (or Lepto- 

 dactylidae), a family of toads, is peculiarly southern. Both in 

 South America and in the Australian-Papuan region, they are 

 the principal elements of the Anuran fauna. Iii addition, there 

 is a single African genus, Helcophryne, which is only known 

 from the Cape and Natal. Its relationships are not very 

 decided, for in some respects it approaches the Avistralian genera 

 and in others the South American. There is no evidence that 

 any of the numerous genera belonging to this family have pre- 

 viously enjoyed an extensive distribution. Each one is confined 

 to a single geographical region, and thus it seems likely that the 

 S;eneric differentiation at any rate has taken place in the southern" 

 hemisphere. 



I have previously expressed the opinion, to which I still 

 adhere, that the known facts are best explained on the assump- 

 tion of former land connections between Australia, South 

 Africa and South America through Antarctica. Bvit Mr. (1. K. 

 Noble," a recent authority, now assures us that there is no 

 need for the Antarctic continent, nor for mid-.-Vtlantic land- 



