14 U.S. NATIONAL MUSEUM BULLETIN 271 



includes the southwestern and southeastern coasts of Australia, South 

 Africa, and Peru to middle Chile; antiboreal includes Tasmania and 

 all of New Zealand, for convenience, and in South America includes 

 all of Schellenberg's (1931) Magellan and Falkland fauna. South 

 Georgian faunas are thrown into antarctic-sub antarctic classification. 

 The tropics of western South America end at approximately 4° S. 

 Warm-temperate of eastern South America is indefinable but of no 

 consequence because of the absence of gammaridean studies along 

 most of that coast except in obvious tropical or cold-temperate regions. 



Gammaridean genera have been found to fall relatively easily into 

 these broad classes, the genera either being confined to one class or 

 being of such wide distribution as to be called cosmopolitan. The 

 latter term, however, primarily refers to genera that radiate outwards 

 from tropical regions into boreal regions but not into arctic-antarctic 

 regions, and there has been little point in so splitting the analysis to 

 segregate cosmopolitan genera that extend high into polar regions. 

 Bathyal, abyssal, and hadal faunas are highly discrete, poorly influ- 

 enced by submergent polar faunas, and, thus, are recognizable as 

 distinct from latitudinal considerations; the few deep-sea genera that 

 have been found only in polar regions have been removed from their 

 endemic position in those regions to the deep-sea classifications. 



No precise statistical methods have been used in dealing with 

 problem genera, those with distributions partially overlapping two 

 classes, because a bit of subjectivism has been applied in each case 

 and because, the principle of "centralism" has been utilized. Genera 

 are thus considered to be confined primarily to that region in which 

 ''most" of the species occur, to wit: a genus with two boreal and one 

 subarctic species is considered to be boreal but a genus with eight 

 boreal, two warm-temperate and one tropical species is thrown into 

 the cosmopolitan class on the probability that more tropical species 

 remain to be described. 



The results of this subjective analysis are presented in table 1; 

 on first sight the data seem to reveal mostly a relationship to study 

 effort, with faunas of low latitudes or southern quartospheres suffering 

 by comparison with well studied boreal-arctic regions. On the other 

 hand the data seem reasonable if one considers that antiboreal regions 

 are few in number, small in size, and low in habitat-diversity com- 

 pared with the extensive boreal regions; except for small oceanic 

 islands the antiboreal region is confined to Tasmania, New Zealand, 

 and two coasts of South America, whereas north boreal regions have 

 four coasts on two continents and a significant disjunct subarctic 

 embayment, the Okhotsk Sea. The rich antarctic shelves are a strong 

 contrast to the polar-arctic impoverishment but perhaps the most 

 striking implication in the data is the low count of tropical endemic 



