I02 TROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



"what occurs. The case of the Chelonian family Pelomedusidce, in 

 whicli the American genus Podocnemis is represented in Madagascar 

 but not in Africa, has already been mentioned. Eesides this, two 

 genera of the typically American Lacertiliau family of Iguanidce 

 occur in Madagascar. It is true that another genus is Polynesian*, 

 so this may be an instance of former wide distribution ; but even in 

 that case, the occurrence in Madagascar of two out of the only three 

 genera that are not American is significant. Amongst Mascarene 

 snakes, four characteristically American genera of Coluhridce are 

 represented — Heterodon (2 species), Liopliis (2), Dromicus (6), and 

 Fhylodryas (2), whilst two other genera are common to America 

 and to other regions. In fact the ophidian fauna of Madagascar, 

 comprising 36 species, is very much more American than African. 

 Lastly, amoDgst the Batrachia, the family Dendrohatidce consists of 

 one genus, Mantella, with 5 species inhabiting the Mascarene Islands, 

 Dendrohates, the only other genus of the family (7 sp.), being Neo- 

 tropical. 



There is one case of affinity between the Neotropical and Ethiopian 

 regions, amongst freshwater shells, that deserves notice. The very 

 peculiar family (Etheriidve, consists of the genus (Etheria, whicli is 

 African, and two genera Midleria and Bartlettia confined to South 

 America. If, however, the West Indies have the same biological 

 relations to the Neotropical region that the Mascarene Islands have 

 to the Ethiopian, and the Lower or Middle Tertiary fauna of each 

 region has been preserved in the corresponding insular groups 

 respectively, the extraordinary development of Cydostomatidce. in 

 the Antilles and in the Mascarene Archipelago may be due to the 

 prevalence of the family in both Africa and South ximerica at 

 the time when they possessed means of communication no longer 

 existing. The occurrence of a few scattered species in Africa and 

 South America, and the extremely poor representation of the family 

 elsewhere, are quite in favour of this view. Even amongst the mam- 

 malia there are some curious relationships ; the only family of In- 

 sectivora belonging to the lower section of the order with narrow 



* This genus is found in the Fiji Islands, which may possibly bare been the 

 eastern extension of the great continent which doubtless at one time included 

 part of Australia, New Caledonia, and JN^ew Zealand. This possibility is ad- 

 mitted by Wallace. The occurrence in the Fiji Islands of three species of frogs, 

 belonging to a genus represented also in the Solomon Islands, New Guinea, tbe 

 Philippines, &c., affords a strong confirmaiion of the view that the Fijis are 

 ancient continental islands. [Mr, Boulenger informs me that a genus of 

 Elapoid snakes {Ogmodoii) is peculiar to the Fiji Islands.] 



