( an ) 



discoverer, did not land on St. Aignan, it cannot be the St. Aiguaii form, and for the 

 reasons given above I do not thiuk that it is Z. pallidipes. 



20. Cinnyris christianae Tristr. 

 This snn-bird, which we have also received from the D'Entrecasteaux IsLands, 

 and from Sudest Ishind, was first described from St. Aignan, where Meek found it 

 tolerably common. 



21. Dicaeum nitidum Tristr. 



This pretty little flowerpecker was first described from Sndest Island, but we 

 have received also a good series from Rossel {a?ited, p. SO), and St. Aignan. " Ou 

 the labels the iris is given as " brown," bill and feet as " black." 



22. CoUocalia esculenta (L.). 



Evidently common, and breeding on St. Aignan. The nests are composed of 

 small grasses and fibres, fixed to the rocks with saliva, and loosely held together 

 by the same material. They are all of a light colour. The eggs measure 15'6 by 

 10-6, 16 by ln-.5, and 16 by lO-O mm. 



23. CoUocalia fuciphaga (Thunb.). 



Tlie brown swiftlet of St. Aignan is no doubt C. faeipkaga, but it must be said 

 that they all have the tarsus quite bare, and that they are rather light below. 

 Specimens from Borneo and Java are generally darker, more smoky on the under- 

 side, but others from the Philippines and the Timor region are just as light on the 

 underparts. 



The nests of C. fuciphaga, are, as a rule, " edible," i.e., consisting of saliva 

 alone. Three nests sent from St. Aignan are not at all " edible," being composed of 

 almost only one kind of long, wiry fibre, of a dark colour, but strongly agglutinated 

 and fastened to the upper surface of some rocks, vis;., resting uii the rock, not 

 hanging on the sides of the rock, as CoUocalia-nests usually do. 



If we suppose that I had made a mistake in uniting (Cat. B. Brit. Mas. 

 XVI. p. 498, and Tierreich, No. 1, p. 67), the C. fuciphagff from the islands of 

 the Malayan and Papuan Islands, then we would be obliged to unite those from the 

 Timor group of islands with those from the Lonisiades, separating them from the 

 darker birds from Java, Borneo, etc., which have generally also feathered tarsi. In 

 the Pliilippine archipelago we find light birds with feathered and with unfeathered 

 tarsi, while Pacific birds are rather dark below, and small. The nests from Borneo, 

 Java, and the Timor group are " edible," or interwoven with feathers and bits of 

 grass, but not entirely built of grass and fibres like those from St. Aignan. Thns 

 the nest-building habits do not coincide with slight differences of colour. It has been 

 said that the same s))ecies of swiftlets build very diff'erent nests, but more evidence 

 is desirable. 



It is possible tliat more subspecies can be distinguished than I admitted in 

 my latest work on the swifts, but more material and exact observations about the 

 nesting-habits are necessary, and no hasty conclusion should be formed on this 

 somewhat difficult subject. 



The eggs from St. Aignan measure 20-5 by 13-.J, 2o-6 by 13-3, 22-7 by 14, 22 

 by 14, l«-7 by 13-7 mm. 



