TERTIARY FLORA OF CHILE 93 



P. robustus Martius of Brazil, is not otherwise known in the 

 fossil state. The genus' contains about 50 existing species 

 confined to the American tropics. The latter, a very doubtful 

 determination, has been compared with the existing A. zns- 

 coidea Poepp. which ranges from Costa Rica to Peru. The 

 genus has few existing species which center in Colombia and 

 eastern Peru. 



The order Ranales is represented by two nominal species 

 of Anona, which, if Engelhardt's figures are reliable, un- 

 doubtedly represent a single species of this'genus. The frag- 

 ment referred to Myristica is valueless as proof of the 

 occurrence of that genus in Chile, but like so many paleo- 

 botanical records it is of geological importance since identical 

 remains occur in the Ecuadorian Andes. 



The family Anonacese is a large one in the modern flora 

 with about two-thirds of its 750 existing species confined to 

 the Old World. The family is evidently an old one despite 

 the paucity of its geological record. The genus Anona with 

 about 70 existing species, nearly all of which are confined 

 to the American tropics, stands in close relationship to this 

 ancient cosmopolitan type. That it originated in America 

 appears reasonably certain from general considerations and 

 from its presence in the Upper Cretaceous of North America. 

 It is the one genus in the family fairly represented in the 

 fossil record, with about a score of species. The Chilean 

 forms have been compared with the existing Anona montana 

 Martius, A. sphcerica Duch., and A. furfuracece St. Hil., all 

 tropical forms of Brazil. The fossils are, however, equally 

 close to various leaves of "chiromoya" observed in Peru and 

 Bolivia, and one is warranted in saying that the "chiro- 

 moya" was a member of the Miocene flora of southern Chile. 



The order Rosales is represented by 7 species all of which 

 belong to the two leguminous families the Caesalpiniaceae 

 and Papilionacese. The family Mimosaceae is unrepresented, 

 in fact the number of leguminous forms known from this 

 flora is surprisingly small and is, I believe, no indication of 



