CHAP. XXX III.] PEYSWAL GEOGIUFUT, 



489 



may posaess representatives of all classes of animals, 

 because many terrestrial Mammals and some reptiles have 

 tlie means of passing over short distaiici^s of sea. But in 

 these cases the number of species that have thus migrated 

 will be very small, and there will be great deficiencies 

 even in birds and flying insects, which we should imagine 

 could easily cross over. The island of Tiuior (as 1 have 

 already shown in Chapter XII L) bears this relation to 

 Australia ; for while it contains several birds and insects 

 of Australian forms, no Australian mammal or reptile is 

 found in it, and a great number of the most abundant and 

 characteristic forms of Australian birds and insects are 

 entirely absent Contraiit this with the British Islands, in 

 which a large proportion of the phmts, insects, reptiles, 

 and Mammalia of the adjacent parts of the continent are 

 fully represented, while there are no remarkable defi- 

 ciencies of extensive groups, such as always occur when 

 there is reason to bidiove there has been no such connexion. 

 The case of Sumati-a, Borneo, and Java, and the Asiatic 

 continent is equally clear ; many large Mammalia, tems- 

 trial birds, and reptiles being common to all, while a large 

 number more are of closely allied forms. Now, geology 

 has taught us that this representation by allied forma in 

 the same locality implies lapse of time, and we therefore 

 infer that in Great Britain, where almost every species 

 is absolutely identical with those on the Continent, the 

 separation has been very recent ; while in Sumatra and 

 Java^ where a considerable number of the eontiiieutal 

 species are represented by allied forms, the separation was 

 more remote. 



From these examples w^e may see how important a 

 supplement to geological evidence is the study of the 

 geographical distribution of animals and phuits, ui deter- 

 mining the former condition of the earth's surface; and 

 bow impossible it is to undei-stand the former without 

 taking the latter into account. The productions of the 

 Aru Islands oiler the strongest e^ddeuce that at no very 

 distant epoch they ibrmed a part of Is^ew Guinea ; and 

 the pecnliaa* physical features whicli 1 have described, 

 indicate that they must have stood at very nearly the 

 same level then aa they do now, having been separated 



