Mr. Wallace on the Geographical Distribution of Birds. 453 



that these two regions are as absolutely distinct as South America 

 and Africa, and it is only because they are separated by straits 

 of from 20 to 100 miles wide, instead of the Atlantic, that they 

 have become slightly connected by the interchange of a few 

 species and genera. 



Thus I account for Gallus reaching Celebes and Sumbawa, for 

 Cervus in the Moluccas, Megapodius in North-western Borneo, 

 a Woodpecker in Celebes, &c. There is, however, an important 

 physical feature which gives us the true key to the separation of 

 the two regions : it is, that the islands of the Indian region are 

 all connected by a shallow sea, while they are separated from the 

 Australian region by an unfathomable ocean. Of this connexion 

 to the Philippines I am not certain, except as far as Palawan, 

 which is joined to Borneo by a 50-fathom bank. Mindanao is 

 also closely connected by islands to Borneo. 



Now look at the map of the Archipelago, and consider that 

 Borneo and Java have species in common by hundreds, Borneo 

 and Celebes only by units, and we shall be forced to believe that 

 the two former have been connected at no very distant epoch, 

 while the two latter have been ever separated, or at least during 

 a long geological epoch, and probably more widely than at pre- 

 sent. Here then is the key to the problem : — Sumatra, Java, 

 Borneo, and the Philippines are parts of Asia broken up at no 

 distant period (an elevation of 50 fathoms would in fact join 

 them all again) ; Celebes, Timor, the Moluccas, New Guinea, 

 and Australia are remnants of a vast Pacific continent in part 

 marked out by coral islands (see Darwin), but broken and sepa- 

 rated at a more distant period, as shown by the fewer species 

 common to the several islands, and the number of distinct sub- 

 faunas into which the region is divided. Celebes is in some 

 respects peculiar, and distinct from both regions, and I am in- 

 clined to think it represents a very ancient land which may have 

 been connected at distant intervals with both regions, or per- 

 haps with some other continent forming a direct connexion with 

 Africa. It may also at one time have had a connexion with the 

 Philippines. All this is indicated by a peculiar genus of Rumi- 

 nants in Celebes {Anoa) ; by a genus of Apes found in Celebes, 

 the Philippines, and Batchian, more nearly allied to the African 



