considerable number of subtropical types are more or less completely 

 separated by the tropical biotas. The phenomenon with which we are 

 dealing is not one of special resemblances between polar elements ^ but 

 rather one of narrow to broad, incomplete to complete equatorial dis- 

 continuities in the latitudinal distribution of plants and animals. It 

 therefore seems desirable that we stop expressing ourselves and thinking 

 in terms of " bipolar distribution " and that we adopt another term — I 

 suggest " antitropical " — to apply to the forms of life that shun the 

 Tropics (the term to apply also to this type of distributicjn^ " Bipolarity " 

 would then become " antitropicality." 



There are man^• patterns of antitropical distribution, because 

 resemblances exist between polar, boreal, cool-temperate, warm- 

 temperate, and even subtropical forms of life ; because the breaks occur 

 along all the main coast-lines that cross the Tropics ; and because the 

 many bitemperate complexes embrace various combinations of two or 

 more temperate biotas (perhaps less than half of the bitemperate series 

 could rightly be classed in the " pantemperate " category, to which I have 

 occasionally referred). These various patterns could be illustrated by 

 different groups of organisms, but I shall largely restrict mv discussion 

 to the marine fishes, with which I am most familiar. 



The bipolar pattern, as already indicated, involves relatively few 

 fishes. In the Antarctic and subantarctic regions there occur, mixed with 

 a much larger number of endemic types, a few species of families that are 

 otherwise chiefly of Arctic or subarctic range — slime-eels (Myxine), 

 skates {Raja), and hakes {Merhtccms) ; also certain cods (Gadidae), 

 sculpins (Cottidse), lump-fishes (Cyclopteridae) , snail-fishes (Liparididae) , 

 seapoachers (Agonidse), and eelpouts (Zoarcidae and Lycodapodidse) 

 (Norman, 1937, 1938). In the far south there are several species of Raja 

 and a moderate number of zoarcids, but the other bipolar groups are 

 each represented there by only one or a very few species. These Antarctic 

 and subantarctic types seem to have been derived from the far north, 

 where the same groups are much better developed. Most, perhaps all, 

 were probably derived by isothermal dispersal, perhaps in part during 

 Pleistocene periods when lesser depths would probably have been required 

 in the transgression of the Tropics. The tropical crossings were probably 

 chiefly effected along the western shores of the two great oceans. Indeed, 

 as a result of the exploration of the United States Fish Commission 

 vessel " Albatross," Garman (1899) disclosed in the region of Panama 

 living deep-sea representatives of half the fish groups here listed as 

 bipolar — namely, Myxine, Raja, Merkiccins, Liparididse, and Zoarcidae. 

 Some of the bipolar fish sets may retain an isothermal connection, as 

 certain invertebrates almost surely do. The varying degree of differenti- 

 ation displayed by the Antarctic representatives of far northern groups 

 • bespeaks widely different periods of isolation. The evidence of relation- 

 ships and of paleichthyology, however, argues against the pre-Tertiary 

 transtropical dispersal of the groups named above. 



Several subantarctic or south-boreal fishes commonly included in 

 the " bipolar " category appear, in contrast, to be ancient, probably 

 pre-Tertiary relicts of groups that have failed to persist in the Tropics, 

 either because of unfavourable physical conditions or because they were 

 unable to compete with the rich faunas that have evolved there. In 

 this category I would include, among fishes, the following fluviatile, 



325 



12 — Pac. Congress 



