DKRIVATIOX OF THE FLORA AND FAUNA 255 



The Myxophvta, 1 8 species, are more or less cosmo|)(jliian [loi). 



I have tried above to indicate where the native plants ha\ e their nearest sta- 

 tions outside the islands or, if endemic, where their closest relatives occur. Sta- 

 tistics like this serve to assign to a local flora its position within a certain floristic 

 region and, when dcalin<^ with an oceanic island, to trace the sources from where 

 its livini; world is likely to haxe been derived. The |)osition j^enerally assif4ned 

 to Juan I'^ernandez is that of an outpost from South America. In 1*Ln'(JLER's Syl- 

 labus the islantls form the "(lebiet von Juan I'^ernandez und Masafuera" under 

 "Zentral- und siidamerikanisches Florenreich"; it is characterized thus: "Gattungen 

 vorzugsweise verwandt mit denen der chilenischen Ubergangsprovinz", i.e. Cen- 

 tral Chile and the Valdivian forest region, T^ngler's transitional belt between the 

 Andean and Magellanian provinces. Goon (/op) distinguished a "Region of Juan 

 Fernandez" under his "Neotropical Kingdom", giving it the same rank as the 

 "Amazone Region", the "Andean Region", etc. We have seen that there is an 

 unmistakable floristic agreement between Juan Fernandez and South Chile, but 

 also that it is far from complete, and both Fngler and Good were well aware 

 of the presence of elements that had little to do with the flora of the mainland. 

 They were barely recognized by Joiiow (see above [). 215) who, with his faith 

 in unlimited transoceanic dispersal, paid little attention to them; they were too 

 few to disturb the Chilean picture. Not until the flora had become better known 

 did its strange features stand out in a clearer light. It is sur{)rising that GlilT-AU- 

 MIX who knew and quoted the synopsis published in vol. II of this work [22^] 

 failed to recognize them. In his paper on the floristic divisions of the Pacific he 

 states that Juan Fernandez lies outside Oceania and he includes it in his discus- 

 sion only for the sake of comparison. The flora is characterized as follows {ji8. 

 931): "Sur 142 Phanerogames indigenes, la moitie sont endemiques mais appa- 

 rentees aux especes chiliennes; les autres sont cosmopolites ou Sudamericaines, 

 surtout chiliennes." It is not easy to understand how he arrived at this conclusion. 

 On the other hand, some authors were led astray by the difference between the 

 islands and the continent. Ermel, who got his imjiression from a short visit to 

 Masatierra, wrote (97.48): 



. . . muss die Zusanimengehorigkeit dieser Insel zum amerikanischen Festlande, zu 

 welcher Ansicht deren geringe geographische Entfernung von selbst hinleiten konnte, aufs 

 nachdriicklichste in Abrede gestellt werden, well die beiderseitige Flora and Fauna zu 

 grosse Verschiedenheiten aufweist, wo wir an geeigneter Stelle die notigen Beweise bei- 

 bringen werden. Die Gestaltung der Flora ordnet die Inselgruppe vielmehr dem australischen 

 Weltteile zu, von dem der grosste Teil heruntergegangen ist. 



It is hardly necessary to mention that his proofs were based on his ignorance 

 of the composition of the island flora and very likely also of the floras of Australia 

 and Chile. 



