318 FOKMKlt LAXI) COXXKCTIOXS. 



Africa were one continent in Cretaceous times, and the con- 

 nection between them persisted until the beginning of the Eocene, 

 these facts would be satisfactorily explained. . . . No known 

 northern fossils can be referred to the African and South American, 

 families, and there is good evidence that the main distribution of 

 fresh- water fishes changed but little during the Tertiary." Mr. 

 Kegan's explanation agrees with that suggested by Mr. Boulenger^ 

 some eleven years previously: — "As it is admitted by most 

 geologists that a continuous land-communication probably existed 

 across the Atlantic between South America and Africa up to the 

 end of the Upper Cretaceous period, it is legitimate to explain 

 the distribution of the Characinidae by such a bridge. This 

 explanation tallies well with the fact, pointing to a severance from 

 remote times, that although the Characinids of the Old and New 

 Worlds show near affinity, no single genus is common to both . ' ' 

 This authority hesitated to explain the Cichlid distribution in the 

 same way, being more inclined to favour a northern origin for 

 the family; still, he wrote: " The hypothesis of a South Atlantic 

 land-communication in the Eocene has much in its favour, and 

 when this is really established all difficulty in explaining the 

 distribution of the Cichlidae will have disappeared." 



Prof. Eigenmann, a leading American authority on South 

 American fishes, has also expressed his opinion that fishes prob- 

 ably interchanged before the beginning of the Tertiary epoch 

 between Africa and South America by way of a land-bridge 

 between Guiana and Africa. He considered that North America 

 has not contributed a single element to the freshwater-fish fauna 

 of South America. 



Evidence derived from the distribution of Arthropods has 

 been epitomised by Mr. E. I. Pococlc" as follows: — 



" Prototiaclieata. — Peripatus is confined to tropical West 

 Africa and tropical Central and South America and the Antilles. 

 Opisthopatus is found only in Chili and Cape Colony [to which I 

 may add also Natal and Lydenburg District, TransvaalJ. 



" DipJopnda. — The bpirostreptid genus Orthoporus, which is 

 of wide distribution in tropical America, is very closely related 

 to tropical African, but not to tropical Asiatic millipedes. 



" Chilopoda. — Parotostigmus occurs in tropical America and 

 Africa, but not in tropical Asia. Scolopendra (S.S.) is mainly 

 tropical and Central American, but in the Old World it has been 

 recorded from the Cameroons, the Canary Islands, Arabia and 

 Socotra. 



" Scorpio)ics. — Of the three tropical American genera of the 

 Scorpionidae, Opisthacanthushas its nearest ally in the tropical 

 and South African Opisthocentrus [the present M'riter regards 

 these two genera as identical] ; and Diplocentrus and Oiclus are 

 closely related to the Arabian and Syrian Nebo, the three together 

 constituting the well marked sub-family Diplocentrinae. 



" Araneac. — The Sicariidae (S.S.) range in America from 

 Chili to Costa Bica, and are only found elsewhere in the world 

 in South Africa. Of the three genera of Caponiidae, Nops and 

 Caponina are tropical American, Caponia South African. 



