DF.RIVATIOX OF THE FLORA AND FAUNA 281 



beyond doubt, and A. a/tescaiu/ois lias its nearest allies in Oceania; comp. above 

 p. 222. 



Aspit'fi/inu. Coi'Ki.AM) ]:)oints out that some groups at least are southern and 

 that perhaps the same applies to the whole genus: "it migrated in its present 

 form from Antarctica" {6g. 167). Of the species found in Juan Fernandez, obiiqnum 

 is austral-bicentric. stcllalum close to neotropical species, inacyosciuin n\o'>\.\\Vit\y 

 of neotropical affinity ; dareoides comes ver\' near ali'arczmsc from Tristan da 

 Cunha and is not, as has been suggested, related to the boreal rt(ta niuraria. 



BlechniiDi. According to CoPKLAND (<5p. 157) no fern genus is more con- 

 spicuously austral in its present distribution and thus more evidently Antarctic 

 in origin. Of the 6 species found in Juan Fernandez aiiriculaiuvi , chiloise and cy- 

 cadifoliu))i belong to an Afro-american group, ScJiottii and perhaps also I'aldiviense 

 of austral-circumpolar type, and longicauda intimately related to the neotropical 

 Sprucei\ even so it cannot blur the picture of an old Antarctic genus. 



Pellaea is an austral-tricentric genus extending north to Canada, but P. chi- 

 le7isis, endemic in Juan F'ernandez, but dangerously close to a widespread American 

 species, belongs to a group that, according to COPELAND [6^.yo), is best allowed 

 to form a separate genus; I have brought it to the neotropical element. 



HvpoUpis is pantropical, especially neotropical, but numerous species are 

 scattered in the equatorial and austral zones: Africa with neighbouring islands, 

 Malaysia and north to Japan, Australia-New Zealand-Lord Howe-Melanesia-Hawaii 

 — when COPELAXD said "without surviving Chilean representatives" he forgot r/^^^- 

 sula. This species is bicentric or, including Folypodiiwi villosoviscidnm of Tristan, 

 tricentric; there is another variety on St. Helena. Thus when COPELAND dj. 177 

 calls Hypolepis "an old antarctic genus now at home chiefly in the Tropics but 

 with two of its paths still occupied", we can add that also the third path is indicated. 



Adiivitnm is most numerous in S. America, and A. cinlense and related species 

 tell us nothing of the earlier history of the genus; CoPELAND 6g. 78: "I suppose 

 that at least the most of the extant species are of Antarctic ancestry, but the 

 genus is old and may have lived elsewhere without interruption"; in 6j. 178 he 

 says that "Antarctica has played a major part in its history", but that the genus 

 "may include elements surviving from other floras and from an age prior to the 

 great dispersal from the south". I refer A. cJiiloise to the neotropical group. 



Pteris. The great wealth of species is in the tropics and only few extend as 

 far south as to south Chile, S. Africa and New Zealand. P. chilcjisis is of neo- 

 tropical character, perhaps also P. semiadnata, while berteroana belongs to a group 

 regarded by CoPELAND as Antarctic to judge from its distribution pattern (see 

 above p. 223). 



Histiopteris mcisa is a pantropical and circum-austral, polymorphous fern with 

 a number of "local derived species" (Copeland) in Indonesia and Polynesia. The 

 genus is, CoPELAND says [6g. 60), "evidently old enough to be a migrant from 

 Antarctica". 



Polypodmm is an aggregation of unities of difi"erent origin and history. Gram- 

 niitis is, in Copeland's opinion (^57. 184), "a plain Antarctic case", and G. Bil- 

 lardieri "surrounds Antarctica more closely and completely than does any other 



