690 



C. SKOTTSBfi;RG 



same relations among the marine algae, with the difference that the independent 

 character of the flora is more pronounced. 



Tiirbcllaria were collected for the first time by our survey (BocK, p. 341). 

 With the exception of a widely distributed pelagic species, they were new (4 sp.). 

 It is of interest that one of them belongs to a genus only known from Japan. 

 Echinoderms (Deichmann, etc., p. 381) seem to be few, a couple of endemic 

 Holothurians, some Asteroids, in part endemic, and 2 Spatangoids. Of the latter 

 one is a subtropical Pacific species, the other was known before only from the 

 Magellanian region and South Georgia. 



Of Cirripeds (NlLSSON-CANTEr.L, p. 483) 4 species have been reported. 

 Two are widely distributed pelagic forms and one of these was found attached 

 to Macrocystis drifted from the continent and associated with a third species in 

 a slightly divergent form, the typical form known only from Talcahuano, south 

 central Chile, on the same substratum. The fourth species ranges from the 

 Strait of Magellan to Peru. 



Finally, NOKDENSTAM (p. 525) has reported on the Tanaidacca and Marine 

 Isopoda. Of the former there are 2 species, both endemic, one with South 

 American, the other with Californian relatives, of the latter 8, 5 of these endemic. 

 Little is said about their systematic position with regard to other species, but 

 one seems to point west, another is very close to a species from the Cape. 

 Of the 3 non-endemic species, i is Atlantic, ranging south to the Cape, and 

 2 are Subantarctic, the nearest locality known being Fuegia or the Falkland 

 Islands. 



From the statements referred to we conclude that the marine fauna of Juan 

 Fernandez is related not only to the fauna of the opposite coast, but also to 

 that part of the South and West Pacific fauna that does not range east to the 

 coast of America. There is a strong endemic element, in many cases not refer- 

 able to the American group. The general character is more Tropical than we 

 have reason to expect from what we know about the mainland along the same 

 latitude. There is also a small but important Subantarctic element which, as a 

 rule, has its most advanced stations at Juan I-^ernandez. These results agree 

 with those obtained through our study of the marine algae. 



Maps of the ocean currents generally allow to the Humboldt or, as it ought 

 to be called, the Peruvian Current (see G. SciiOTT in Annalen der Hydrographie 

 59. 193 1. P- 161) a width sufficient to embrace the Juan Fernandez as well as 

 the Desventuradas Islands. Several of the zoological authors quoted above have 

 pointed out that this position explains the presence of cold water forms — in 

 cases even called cryophilous , which is, I think, to go a little too far — at 

 the islands as well as on the mainland coast, and they do not seem to pay much 

 attention to Plate's statement that Juan P^ernandez lies outside the current. 

 Plate tried to explain the faunistic difference between Juan Fernandez and the 

 mainland as a result of difference in temperature. He quotes (1. c. p. 226) 

 Bergiiaus' Atlas der Hydrographie, where the Peruvian Current is drawn so 

 narrow that it does not reach the islands, and KuMMEl/s Handbuch der Oceano- 

 graphie which states that its width is only about 100 miles; farther offshore 



