ANNIVEESAKY aKDEESS OE THE PEESIDENT. lOI 



is almost equally divided between South America and Africa, except 

 that one genus extends north into the Mediterranean area, and that 

 two are found in North America. That the original connexion was 

 between the southern continents is, however, indicated by two of the 

 genera, Amphishcena and Anops, being represented in both, whilst 

 the Palaearctic and Nearctic genera are not nearly related to each 

 other. Another Lacertilian family peculiar to Africa and South 

 America is that of the Anelytropidce ; but these are not of much 

 importance. Some genera of snakes, e. g. Ahcetulla, Dryiojpliis, 

 Dvpsadoboa, and Leptodira, have the same distribution. One genus 

 of apodous Batrachians (Cceciliidce), Dermpphis, is also African and 

 S. American. Amongst the Batracliia the most remarkable 

 instance, however, is afforded by the Ar/lossa, a low but peculiarly 

 specialized group of toads, of which one family, Pipidce, is purely 

 Neotropical, the other, Xenopidce (Dactijlethridce), peculiar to Africa. 

 It is possible that this may be a case of a group having formerly a 

 much wider range, and the same may be the case with the dipnoans 

 Lepidosiren and Protojpterus ; but it is difficult to account for the 

 distribution of the Ghromididce, Gharacinidce^ and AmpJiishcenidce on 

 such an hypothesis. All three are well-marked and well-developed 

 families in both areas. Of the Chromididce, fossil representatives, 

 especially Pycnosterinoc^ are said to have been found in the Cretaceous 

 rocks of the Lebanon ; but the relationship of these forms is not free 

 from doubt, and in any case they were marine, and the marine repre- 

 sentatives of the Chrornididce are extinct. It is also true that a 

 wide-spread marine family, the Lahridce, is closely allied to the Chro-- 

 mididce, so that although it is far from probable that the African mem- 

 bers of the latter have originated separately from the American, such a 

 contingency might be suggested but for the evidence afforded by the 

 Characinidce. This family is unknown in the fossil state, and there 

 is nothing to indicate that it ever inhabited Europe or North America. 

 There is another piece of evidence. If Africa was formerly in 

 land communication with South America, it is probable that before 

 the Ethiopian fauna was profoundly modified by the incursion of 

 Palaearctic types in the Pliocene and Pleistocene periods, several Neo- 

 tropical forms that are wanting there at the present day existed in 

 South Africa. If this was the case, and if, as has already been 

 pointed out, there is a remnant of the old African fauna, preserved 

 from contact with the improved Palaearctic forms, in Madagascar, 

 several alliances with S. American types should be found there that 

 can no longer be traced in continental Africa. This is precisely 



