MAKINK AI.GAI. COMMUMI IKS OI' 1 UK JLAN KKKNANDKZ ISLANDS 693 



south. In order to tix its southern extension we have to lo(;k for the most 

 soutlierl)- upweUing centre. Sciloi I found that upwelhng began just north of 

 29°, whereas GL"N'rilP:i< sa>'s about 40 , which seems more hkel)-. This boundary 

 has, in my opinion, a direct bearing on the distribution of the Httoral flora 

 antl fauna. 



The juan Fernandez and Desventuradas (San Felix and San Ambrosio) 

 Islands rise precipitously from the Juan F'ernandez ridge. They are neovolcanic. 

 According the opinion of most geologists and some biologists they are truly 

 oceanic and ha\e ne\er been connected with an\' other land. I'^ew geologists 

 admit the possibility of land connections in this region, whereas a number of 

 biologists have declared themselves unable to understand their organic world 

 unless suitable connections existed in former times; they regard them as crypto- 

 oceanic, representing the highest volcanic peaks of a submerged land, which 

 may have formed a westward extension of South America. Both flora and fauna 

 are rich in peculiar endemic genera and species which have no relatives in South 

 America and which cannot have evolved from stray oversea immigrants since 

 the volcanic islands, such as they appear to-day, were formed. A considerable 

 proportion of the endemic plants has been referred to an Old Pacific element, 

 connected with the Antarcto-Tertiary flora and by way of Antarctica with Xew 

 Zealand. This ancient element is also represented in South America by single, 

 isolated forms. Other insular endemics point towards the Tropical Andes. The 

 upheavel of the Andes, a gigantic process, revolutionary also from a climatic 

 standpoint, the formation of the new coastline, the submergence of the Juan 

 Fernandez land under volcanic eruptions gradually giving rise to the present 

 islands, seem together to offer an explanation for the accumulation of endemic 

 types and for the difference between the organic world of the islands and that 

 of the continent, a difference accentuated during the Pleistocene glaciation, 

 which had but little effect on the islands. I have discussed this question on 

 several occasions; see for instance K. Svenska Vetensk.-akad. Handlingar 51 

 (1914) and B. P. Bishop ^Museum Bull. 16 (1925). 



Land connections seem to be of little consequence for the distribution of 

 the marine fauna and flora. Plate (1. c. p. 228) is convinced that Juan Fer- 

 nandez rose from the ocean floor to remain permanently isolated; the fauna, he 

 says, would have been predominantly Chilean, had there ever been a land con- 

 nection. He thinks that it immigrated across the sea from Peru and North Chile; 

 species not living there now may have existed there before. The flora, then, 

 should have the same origin. Our knowledge of the dispersal of marine algae 

 is very limited. Not a few are distributed over large parts of the globe, and 

 seem to be able to cover large distances and to establish themselves under rather 

 different conditions; they are eurythermic with a wide amplitude and have be- 

 come almost cosmopolitan. Some species are known to have extended their 

 area considerably during the last few decades, at least along a coast. On the 

 other hand, many are known from a few places separated by very large distances 

 and not known from any intermediate station. In many cases the discontinuous 

 distribution is apparent, and the gap will be filled by future research work, in 



