334 MR. A. R. WALLACE ON THE BIRDS [DeC. 9, 



lieving such to be the case, I should probably have taken no trouble 

 to obtain a collection from thence, had I not been told by many of 

 the natives who trade to Sula that a beautiful little bird of the Par- 

 rot family was found there and in no other place. In consequence 

 of this and other more or less vague information about its productions, 

 I arranged with my assistant, Mr. Allen, to go there for two months. 

 Owing to bad weather, ill health, and the usual troubles about boats, 

 men, and provisions, he obtained but a very small collection, made on 

 the southern and eastern islands. Only forty-eight species of birds 

 were obtained, yet out of these there were seven new species, which 

 appear to be altogether peculiar to this little group of islands ; five 

 or six others are rare birds of the Moluccas or Celebes, and the re- 

 mainder the commoner species from the same countries. 



But although the Sula Islands show a mixture of the forms of 

 Celebes and the Moluccas, yet these countries have not contributed 

 towards its fauna in anything like an equal proportion. Deducting 

 ten species which have a wide range over a large portion of the Ar- 

 chipelago, and even beyond it, and dividing the remainder into two 

 portions — those that may be supposed to have been derived from 

 Celebes on the one hand, and from the Moluccas and islands to the 

 east and south of them on the other, — we shall find that the Celebesian 

 forms are almost exactly double the rest. Twenty species are iden- 

 tical with birds found in Celebes, and five new species are of Cele- 

 besian forms ; whereas only eleven species are found also in the Mo- 

 luccas, and but two of the new species can be affiliated to Moluccan 

 types. Twenty-five of the species of the Sula Islands must there- 

 fore have been derived from Celebes, and only thirteen from the 

 Moluccas. The accompanying Table (p. 335) shows the species 

 distributed according to their derivation. 



It is further interesting to remark that all the Raptores and all 

 the Pigeons and Parrots, but one of each group, are Celebesian species 

 or forms ; while among the Moluccan species are many active but 

 weak-flying birds, including five species of Flycatchers, which would 

 be most likely to be carried over by strong winds. Further, the 

 birds derived from the Moluccas contain three genera which do not 

 occur in Celebes. 



From these facts it seems to me clear that the Sula Islands are 

 really an outlying portion of Celebes, and must at some former period 

 have had a much closer connexion with that great island than at 

 present. The Moluccan species must therefore be considered as im- 

 migrants, many of them from Bouru, which is only forty miles di- 

 stant ; and the fact that some of these early Moluccan immigrants 

 have already become modified into distinct forms, some of which may 

 be classed as species, others as varieties, shows for how long a period 

 of time the small and scattered islands of the Moluccas must have 

 remained in their present disconnected state. 



The following Table shows the geographical affinities of the birds 

 of the Snla Islands : — 



