372 C. SKOTTSBERG 



done to examine the <^enoty})ical constitution of this astonishing assemblage of 

 \oca\ forms; some may, for ail we know, be ecotypes, but experimental studies 

 have not been made. A limited number of hybrids have been described, but we 

 know nothing of their behaviour. Anyhow, the floristic difference between the islands 

 is very striking and hardly in favour of the theory of transoceanic migration, for 

 tlie majority t)f organisms seem to have been unable to cross even the straits 

 separating the islands, and biotic factors preventing establishment can hardly be 

 made responsible in so many cases. 



Chapter IX. 

 Juan Fernandez — oceanic or continental ? 



To W'ai.i.ack as well as to the majority of biogeographers the Juan Fer- 

 nandez Islands were typically oceanic in spite of their moderate distance from 

 South America. They had the advantage of antiquity, WALLACE remarks, for 

 the means of transmission had formerly been greater than now, their surface was 

 varied, soil and climate favourable, "offering many chances for the preservation 

 and increase of whatever plants and animals had chanced to reach them" (p. 287), 

 I lad the character of Masafuera been known to him he might have been less 

 optimistic. The land-shell fauna, entirely endemic, testified to the great age of 

 the islands, for none had been introduced for so long a period that all which 

 did come had given rise to new forms — or were the last of a fauna extinct 

 on the continent. 



Joiiow (/jo) based his opinion on Wallace; when discussing the dispersal 

 agencies and the morphology of the diaspores his starting-point was /wo islands, 

 Masatierra })lus Santa Clara and Masafuera, which has risen separately from the 

 deef) sea. It is strange that he never thought of another possibility, because the 

 geologist who went with him and who wrote a chapter on the geology of Masa- 

 tierra, claimed to have discovered, in one place, a fundament of rocks older than 

 the omnij)rcsent young basalt — ^JoilOW could not know that the interpretation 

 of this stratum was false, as later shown by OUENSEL (J02). After his visit to 

 the ncs\ enturadas Joiiow modified his opinion; the submarine ridge uniting these 

 little islands with Juan T'ernandez had then been discovered: 



Die iiiUer [.'Icicher P.reite mit dem Hafen Caldera und in derselben Entfernung 

 vom Kontinent wic Juan I'crnandc/. ^elegene Inselgruppe ist vulkanischen Ursprungs 

 und stcilt, wie die von dcin Mitglicd der Expedition, Herrn Chaigncau, ausgeliihrten 

 Eotungen crgahcn, die iiher Wasscr befindiichen hochsten (iipfel einer im Ubrigen 

 iinterseeisch verlaufenden liergkcttc dar, welcher auch die Inseln der Juan-Fernandez- 

 (iru])pe als siidli^ hstc (iiptel angehoren. Aus dem Vergleiche der Eloren und Faunen 

 heider .\r( hipelc, \vcl( he trot/, der grossen klimatischen Verschiedenheiten frappante 

 X'crwandtschaft autweisen, crgicbt .sich mit zwingender Notwendigkeit die Hypothese, 

 dass die zwei Inselgriii)i)en in der Vorzeit mit einander in Landverbindung gestanden 

 hahen und dass ihre Isolierung die Folge einer stattgehabten Senkung jener Bergkette 

 ist isso- 259). 



