316 



SCIENCE. 



[N. S. Vol. III. No. 61. 



have been some connection between Africa 

 and South America early in tiie tertiary 

 epoch. This connection in the present con- 

 dition of our ignorance of paleontological 

 facts, appears to be more probable than the 

 derivation of the common peculiarities of 

 the faunas of the two continents from a 

 former cosmopolitan fauna or northern areas 

 which have lost them, leaving them to the 

 two southern continents only. The union 

 of Africa with Asia culminated too late 

 to allow of much differentiation of the 

 invading forces that spead over its wide 

 domain. 



A former quasi-cosmopolitan fauna was 

 nevertheless manifest in the case of the 

 Ceratodontids, but in Europe and ISTorth 

 Amei'ica they flourished early in Mesozoic 

 times, and none survived later than the 

 Jurassic, and approximately coeval with 

 them were species which lived in India and 

 Africa, but all these died out and the only 

 survivors are the species of Neoeeratodus of 

 tropical Australia. This family was men- 

 tioned as an extreme case of persistence for 

 an osseous fish type. 



The amphibians furnish very ambiguous 

 evidence if the accepted taxomy is correct. 

 For example, on the one hand the Cysti- 

 gnathids are well developed and limited to 

 America and Australia, but on the other the 

 Discoglossids are all European, except one 

 genus (Liopelma), and that is confined to 

 New Zealand. 



The reptiles contribute data looking in 

 different directions. One of the ablest 

 herpetologists of all time has expressed the 

 opinion that ' if a division of the world had 

 to be framed according to the lizard fau- 

 nas,' the Ethiopian and Palfearctic regions 

 should be combined in one ( Occidental') and 

 the Australian and oriental in another ( Ori- 

 ental) , to be themselves aggregated in a realm 

 (PALiEOGEAN) differentiated from another 

 (Neogean) , comprising the Neotropical and 

 Nearctic regions. Their mode of distribu- 



tion in fact approximates that of birds, but 

 has been seriously affected bj^ their intoler- 

 ance of cold and consequently the loss of 

 types, which might be interchanged between 

 the continents. The similarity between the 

 African and Palsearctic regions is doubtless 

 due to the intrusion of forms from the 

 latter into the former. The African, how- 

 ever, has three small families restricted to 

 its area and two shared with America. 

 Quite different is the distribution of the 

 tortoises. 



The superfamily of the Pleurodirous or 

 Chelyoidean tortoises is restricted to the 

 southern continents. One family (Ster- 

 nothserids) is peculiar to Africa, one 

 (Chelyidse to America and one (Chelo- 

 dinids) to the Australian realm, while one 

 (Podocnemidids) is common to Africa and 

 America, and another (Ehinemydids) to 

 America and Australia. Except in America 

 these completely replace the fresh water 

 cryptodirous tortoises, but it is noteworthy 

 that species of the terrestrial Testudinids, 

 generally considered as congeneric, occur 

 in all the warm continents except the Aus- 

 tralian. It must not be forgotten that 

 formerly (in early tertiary times) the Chely- 

 oideans were represented and, it has been 

 claimed, even by a still existing genus 

 (Podocnemis) in the noi'thern hemisphere, 

 and therefore their present occurrence only 

 in the southern continents loses much of its 

 significance. The evidence of former con- 

 nections of the southern hemisphere fur- 

 nished by both amphibians and reptiles is 

 indeed of very little account per se and is 

 only significant as collateral to that pre- 

 sented by Qther classes. 



To sum up the results of studies of the 

 several classes, the present evidence points 

 to a comparatively recent union of or con- 

 nection between the southern continents. 

 The inference (independent of the ichthyo- 

 logical data) is based in part on the infor- 

 mation respecting the geological duration of 



