276 



C. SKOTTSBERG 



during late Jurassic and very early Cretaceous times through three "gates" in three 

 main streams, the West Polynesian, the Magellanian and the African gate. Many 

 of the genera occurring in Juan l>>rnandez, endemic or not, are mentioned by Croizat ; 

 several of them are, also in my opinion, of Antarctic origin. 



Onoholus. Instead of allowing it to enter S. America through the "Magellanian 

 gate ' Ckoiza r makes a circuit along his Pacific baseline from New Zealand to F'uegia. 



Juauia. I ha\ e already discussed the taxonomic position of this genus which 

 Ckoi/ai' derives from a Mascarene centre in company with related Andine genera. 

 Ik nillNSON linked it with Ceruxyloii to form a tribe Ceroxyleae; Croizat remarks 

 that 



this mav be better taxonomy, we do not know, but does not change substances in the 

 least, 'i'he ingredients that make up the Morenieae, Iriarteae and Ceroxyleae are all from 

 the same kitchen. In one case Hyophorbe speaks, in another so does the face of dispersal. 

 Kvervthing in the end rhymes to the very same. With or without Hyophorbe, seasoned 

 with this or that sauce, the stew does not change . . . classification of such a genus as 

 this is bound forever to remain the j)laything of opinion (pp. 456-457), 



a statement quite characteristic of this author's method of approaching a phyto- 

 geogra[)hical problem. 



Pipcyo))iia has its stronghold in tropical America where a number of subgenera 

 are found. To CROIZAT the genus is Pacific and Juan F'ernandez one of the starting- 

 points of its dispersal toward the heart of S. America, the four island species leading 

 straight to north Chile, lie based this opinion on YUNCKER who wants to bring 

 them to subgen. SphacrocarpidiuDi^ where, however, only one of them belongs, while 

 the other three, as was brought forth above (p. 259), point west, one of them inti- 

 mately related to P. ty'istanciisis, which, unexpectedly it is true, brings the Antarctic 

 within sight. In a footnote CkOlZA T admits that I may be right in my opinion of 

 the taxonomic {)osition of P. bcrtcroana\ none the less he finds the contact Juan 

 hYM'nandez- Tristan not at all j)eculiar. EDipetruni is another example, both "a 

 mere asj)cct f)f the standard dispersal between Africa and South America" (p. 102). 

 Hut tiiis "standard dispersal", as discussed by Engler, refers to a number of 

 tr()j)ical genera and families bearing witness of a transatlantic bridge. In Croizat's 

 opinion the track Mascarenes-Africa-America was a very important route. 



I'alms and Pepcromias are pantropical and range south to New Zealand and 

 central Chile, and there are man\' endemic palm genera on the Indian and Pacific 

 islands south of the Ivpiator, whence follows that the field is open for speculation 

 and that the possibilit)- of a ])rimary centre in the south may be discussed without 

 breaking the rules of the game, but to derive exclusively neotropical families from 

 a Mascarene centre cannot fail to cause astonishment. Croizat, dealing with Cac- 

 taceae, Bromeliaceae, Kai)ateaceae etc. — and in an earlier chapter he asserts that 

 Malesherbiaceae also had their starting j)oint in Africa — says that 



to credit grouj)s of the kind with "American" origin steps must be taken as a prelim- 

 inary which no one trained in |)rccise thinking will encourage. . . . Rhipsalis is doubtless 

 a (iondwanic genus because it is rooted within the triangle Ceylon-Madagascar-Masca- 

 renes. It well may have reached ]')razil and the West Indies from West Africa. . . . the 

 baseline of the Cactaceae scpiarely rests between Patagonia and the Mascarenes, and if 



