210 



NATURAL HISTORY [ciiap. iiv 



islands which have never formed pai't of the adjacent 

 lands, although so closely approacliitig tlieui ; aud their 

 productions have the characteristics of true Oceanic Islands 

 sUghtiy modified. These characteristics are, the absence 

 of all Mammalia except bats, and the occiurence of 

 peculiar species of birds, insects, and land shells, which, 

 though found nowhere else, are plainly related to those 

 of the nearest land. Thus, w^e have an entire absence of 

 Australian injimmals, and the presence of only a few strag- 

 glers from the we-st, which can be accounted for in the 

 manner already indicated. Bats are toleiably abimdant 

 Birds have many peculiar species, with a decided relation- 

 ship to those of the two nearest masses of hind. The 

 insects have similar relations with the birds. As an ex- 

 ample, foui' species of tlie Papilionidie are peculiar to 

 Timor, three othera are also found in Java, and one in 

 Australia. Of the four peculiar species two are decided 

 modUications of JavMiese forms, wlule the others seem 

 allied to those of the Moluccas and Celebes. The very few 

 laud shells known are all, curiously enough, allied to or 

 identical with Moluccan or Celebes forms. The Tieridai 

 (wliite and yjeUow butterflies) which wander more, and 

 from frequenting open grounds are more liable to be blown 

 out to sea, seem about equally related to those of Java, 

 Australia, and the Moluccas. 



It has been objected to Mr. Darwin^s tlieor)',— of Oceanic 

 Islands having never been connected with the mainland, — 

 that this would imply that their animal population was a 

 nutter of chance; it has been termed the ''Jloimm and 

 jetsam theory*' and it lias been maintained Umt nature 

 does not work by the " chapter of accidents," Eut in the 

 case which 1 have here described, we have the most posi- 

 tive evidence that such ha^ been the mode of peopling the 

 islands. Their productions ar^ of that miscellaneous cba- 

 raoter which we should expect from such an origin; and 

 to suppose that they have been portions of Australia or of 

 Java will introduce perfectly gratuitous difticulties, and 

 render it quite impossible to explain those curious rela- 

 tions which the best known group of animds (the birds) 

 have been shown to exhibit. On the other hand, the 

 depth of the sm-rounding seas, the fonn of the submerged 



