FORMER LAND CONNECTIONS. 829 



bridges, thus:—" If we assume a (north) polar origin for the 

 Pipidae, Hylidae, and Leptodactyhdae, we escape the necessity 

 for building any land-bridge." Tlius we have to assume tliat 

 there was once a Cystignathid fauna throughout the northern 

 Jiemisphere and throughout Africa, which fauna has completely 

 disappeared except for the southern species. There is no actual 

 evidence in favour of this : it is offered merely as a simpler 

 assumption than that of the objectionable land-bridge. 



From the work of Mr. Noble and several earlier authorities, 

 we realise that the family Cystignathidae is not really separable 

 from the Bufonidae : these latter seem to be merely toothless 

 Cystignathids. The toothed Cystignathid genera are to be 

 regarded as primitive, and the various Bufonid genera as derived 

 thei'efrom. The position may now be stated, as follows (excluding 

 from consideration the genus Bufo, which is a recent and hardy 

 cosmopolitan, except in Australia) : 



(a) There are 30 genera, some toothed and others toothless 



in South America, and two additional genera restricted 

 to Central America, but except for one or two stragglers 

 in Northern Mexico and Florida none of these extend 

 their range beyond the neotropics, an*! of those genera 

 which occur north of the isthmus " probably the majority 

 have pushed their way northwards since the Panama 

 connection " (G. K. Noble). 



(b) There are 12 toothed genera and three toothless in. 

 Australia : two of the former and one of the latter occur 

 in Tasmania : New Guinea has two toothed genera : 

 the Indo-Oriental region has no toothed genera whatever, 

 but has a number of toothless ones. 



(c) Africa has a single peculiar toothed genus in the extreme 



south, and several toothless ones in the tropics, whicli 

 latter, however, are not peculiar, but are represented 

 also in the Oriental region and Australia : one of them 

 (Pseudophryne) is derived directly from an Australian 

 toothed genus and the other is hai'dly separable from it 

 according to Mr. Noble. Madagascar has no representa- 

 tives whatever. 

 One conclusion that may be lawfully drawn front these facts 

 is that the toothless Bufonids now living in tropical Africa and 

 some, if not all, of those in the Oriental region, have migrated 

 there as such from the headquarters of the Cystie;nathidae in 

 Australia: thus, they do not provide a missing link in the 

 hypothetical world-wide distribution of the Cj'stignathidae. 



A consistent theory niust embrace also all other members 

 of the sub-order Procoela, such as the families Hylidae and 

 Brachycephahdae. Concerning the latter, nothing more need be 

 said than that it is confined to the neotropical region, where no 

 doubt it originated. The Hylidae, from structural considerations, 

 have to be regarded as derivations of the Cystignathidae. The 

 main features of their distribution are as follows: — Abundant 

 in the South Amei'ican and Australian regions ; completelv absent 

 from the Indo-Oriental region and from tropical and South 



