2 EINAR LÖNNBERG, (Schwed. Südpolar-Exp. 



The collection comprises as well shore fishes as pelagic, resp. benthopelagic, 

 fishes. The latter will be described in a separate chapter. The former again are 

 collected at many difi"erent stations as well in the subantarctic as within the true 

 Antarctic Region. The localities group themselves, however, naturally round certain 

 geographical areas, viz. Tierra del Fuego with Staaten Island and surrounding seas, 

 the Falklands with the Burdwood Bank, South Georgia, and finalh' the South Shet- 

 lands-Graham Land complex of islands and lands. I have therefore found it most 

 suitable to treat the fishes of these areas separately, the more so as, as will be shown 

 in the following, these areas from a zoogeographical point of view, to a certain de- 

 gree, form units. By this I mean that the fishes of one such area are not all of them 

 wholly identical with those of another area, but at least some of them represented 

 by similar fishes which in certain instances, although in many respects corresponding, 

 are specifically difterent, in others only subspecifically, or racially. This difference 

 is a natural result of isolation, because the shore fishes of one district have been 

 prohibited by wide interjacent areas of deep water to interbreed with their con- 

 geners in another district. This can, of course, only hold good for such fishes which 

 have demersal eggs, and which, at no period of their life, lead a pelagic life. Al- 

 though the development and life-history of the Nototheniidœ are very imperfectly 

 known, it might be assumed per analogiam from what we know about arctic fishes, 

 that shore fishes living in such a cold climate, as most of the Nototheniidae do, 

 hardly can have a pelagic development. In certain in.stances the comparatively large 

 size of the eggs indicate that they are demersal. It also happens, especially among 

 the members of this famih-, that geographical species, resp. subspecies, have been 

 developed and substitute each other within different districts. 



When the differences between the representative species are very great and the 

 characteristics easily seen no systematist would hesitate to describe each under a 

 separate name. When the distinguishing characteristics between the fishes of one 

 region and those of another are less sharply marked and less numerous, the opinion 

 of different ichthyologists might perhaps take different expressions. Some might 

 create new species, others might unite several forms under one and the same name. 

 It is the old case "splitter" versus "lumper". It seems to me that both extremes 

 should be avoided. When two fishes from two different localities are essentially 

 alike and exhibit the same type, so to say, but at the same time differ through 

 some perhaps small, but always constant characteristics, they may be regarded as 

 belonging to one and the same species, but as being subspecifically distinct, 

 and to express this, three names may be used, as rather extensively has already 

 been done in the masthological and ornithological literature, but comparatively little 

 in the ichthyological. In such a way the unity as well as the diversity has been 

 duly recognized, as they ought to be. When describing the present material I have 



