1863.] MR. A. R.WALLACE ON THE GENUS COLLOCALIA. 383 



east, and west of Amboyna, where Rumphius obtained his specimens. 

 It is to be observed that these white spots are quite concealed, both 

 on the upper and under view, by the overlapping of the feathers and 

 the tail-coverts ; so that, as Rumphius accurately describes it, " only 

 when the feathers are separated, the white spots become visible." 

 This circumstance, and the close general resemblance of the bird to 

 the allied CoUocalia linchi, Moore, which has no tail-spots, has led 

 to this striking character being overlooked. It would appear that 

 these two species are restricted to the Australian and Indian regions 

 of the archipelago respectively, C. linchi extending from Java west- 

 ward to the Nicobar Islands, while C. esculenta is found in Celebes 

 and through the Moluccas to Timor and the shores of New 

 Guinea. 



It seems extraordinary that a bird which ranges over nearly half 

 of the Malay Archipelago, and is by no means uncommon there, 

 should not have been hitherto identified with the Linnaean species, 

 as it is almost certain that specimens of it must exist in the museums 

 of Leyden and other Continental cities. The fact may, however, 

 probably be accounted for by the circumstance of one of the autho- 

 rities most relied upon having been himself deceived in what he sup- 

 posed to be the constructor of the edible nest of Java. M. Poivre 

 furnished Buflfon with a figure and description of a bird obtained at 

 the place where nests were found on an island in the Straits of Sunda. 

 This figure and description have been copied by Brisson, and have 

 been mixed up with the description of the true esculenta by Grnelin 

 and other naturalists. The locality, however, from which M. Poivre 

 obtained his nests would show that they must have belonged to C. 

 linchi or C.fuciphaga, and not to C. esculenta ; while his figure 

 proves that the bird he obtained was not a CoUocalia at all — not 

 even a Swift or a Swallow, but a short-winged, long-legged, straight- 

 billed bird, agreeing in du^ky colouring with CoUocalia fuciphaga, 

 but having the tail-feathers white-tipped. These extraordinary dis- 

 crepancies, however, seem to have been generally overlooked ; and the 

 character given by Brisson, " rectricihus nigricantibus, apice albis" 

 has been taken as agreeing with that of Linnaeus, " rectricihus om- 

 nibus macula alba notatis ; " whereas Rumphius (who is quoted by 

 Linnaeus) says that the feathers must be separated in order to see 

 the white spots, clearly proving that they were situated towards the 

 base, and not at the tip of the tail*. 



In the British Museum collection there is a very beautiful species 

 from the New Hebrides, having the white tail-spots as in C. esculenta, 

 but with a narrow band of pure white across the rump. 



I add a synopsis of what appear to me to be the well-authenticated 

 species of this genus hitherto described. 



* Since reading this paper, I have found that the late Prince Bonaparte had 

 made the discovery of Hirundo esculenta, L. (Coroptes Uendus, xli. p. 977), in 

 1855, from a Timor specimen in the Paris Museum, collected by Mauge in 1820. 

 He notices the white spots towards the base of the tail. With this independent 

 confirmation of my opinion, there can remain no doubt that the species is now 

 satisfactorily determined. 



