160 BARBOUR: ZOOGEOGRAPHY. 



as one would expect, British New Guinea — i. e. that part of the island l3dng 

 along the Papuan Gulf and Torres Strait — supports the great majority of the 

 forms which are so strikingly similar to those of Queensland. The western end 

 of New Guinea, even now almost cut off from the main bulk of the island by the 

 Great Geelvink Bay and McCluer Gulf, has a very distinct fauna, and certainly 

 has not always been connected with the rest of Papua. The birds of the region 

 of Mt. Arfak are almost always specifically, and often generically, distinct from 

 their relatives in the Owen Stanley and Finisterre ranges. The same thing is 

 almost equally true of the ornis of the coastal plain region. The reptiles and 

 amphibians are strikingly dissimilar, as a glance at the table of distribution 

 will show. Again, the question is not yet conclusively settled as to whether 

 Celebes was pushed up from the sea as a single island with shape similar to that 

 which it now has; or whether, as Weber holds, it has been formed by the con- 

 solidation of several separate islands, which have each received their fauna from 

 a separate region, and these islands, having fused, gave Celebes the composite 

 fauna which it now supports. This particular question, however, is really of no 

 special importance. The point of real importance is the three-fold origin of the 

 fauna itself. 



The relation of the Philippines to Halmahera is a question which still awaits 

 solution. It seems possible to project the line of recent volcanoes through 

 Halmahera up to Mindanao; in which case land may well have existed along a 

 similar line. Lines of recent extensive faulting often give rise to volcanoes, and 

 this rnay have, been the case here. Such a connection, however, can hardly be 

 urged as a substitute for the Celebes-Halmahera Bridge. The types which 

 suggest immigration from Celebes do not occur among the southern Philippines, 

 except for some on Palawan. The relation of Mindoro to Celebes, suggested at 

 once by the distribution of the pigmy buffaloes, is, according to Bartsch, also 

 evident from a study of the land-snails. Mindanao, with a fauna different from 

 that of Celebes and Mindoro, must needs be of more recent origin. It has 

 probably replaced, by having been lifted again from the sea, some of the land 

 which became submarine between Celebes and Mindoro; and, joining with 

 other islands, received a typical Malayan fauna from Borneo, and some Celebe- 

 sian types from small islands that may have represented unsubmerged mountain 

 peaks of the older land-mass, and that supported some of the types common to 

 Mindoro and Celebes. The Papuan element in the amphibian fauna of Borneo 

 may be a true relic-fauna; for the engystomatids, which exhibit such a very 

 noteworthy elaboration in Papua, may have come from Borneo to Halmahera 



