THREE UNPUBLISHED PAPERS BY ELYTH. 315 



uniform, differing onlj^ in the colour, according to the relative age 

 of the nests. It exhibits none of those diversities which might 

 be expected if it were collected casually (like the mud employed 

 by Hirundo urbica, and the materials commonly employed in 

 nest-making), and applied to the rocks. If it consisted of the 

 substances usually supposed it would be putrescent and diver- 

 sified." 



" Dr. Horsfield," continues Sir S. Raffles, " thinks that it is 

 an animal elaboration, perhaps a kind of secretion ;"* and Sir 

 Everard Home, after examining the stomach of one of these 

 birds, pronounced his decided opinion that the edible nests are 

 composed of a substance secreted by the glands of that organ.t 

 M. Poivre, who in his often-quoted communication to Mont- 

 beillai-d (Buffon's Hist. Nat. Ois. vi. 688, et seq.), advocated the 

 exploded notion that the substance was collected from the sea, 

 remarks that there have been seen threads of this viscous matter 

 hanging from the bills of these birds ; and that it is believed, 

 though without any foundation, they were derived from the 

 stomach during the nesting season. M. Hooyman long ago 

 (1781) was of opinion that the substance of them was nothing to 

 do with sea-foam, but is elaborated from the food of the bird.+ 

 The Rev. J. Barbe adds, to his account of the nests gathered in 

 the Nicobars, &c., quoted in the sequel, that " The Chinese 

 say that when the nest is taken before it is completed the bird 

 makes another, but of an inferior quality ; and it appears that it 

 exhausts itself in building the second, the nest being spotted with 

 blood." § The prevalent opinion is that the substance is secreted 

 by the salivary glands ; but though, on carefully dissecting a 

 sjDecimen of C. fucii^haga, killed when there is reason to supjDose 

 that it had young (and therefore too late in the season), no 

 remarkable development either of the salivary or of the proven- 

 tricular glands was apparent, yet we are fully convinced that it is 

 secreted by the latter, from examination of other Cypselidce ; even 

 in Cyjiselus balasiensis, which emj^loys so little mucus, which is 

 laid on merely to make its nest adhere, we have detected in 

 situ stringy clots of it, while the proveutricular glands were 



* ' History of Java,' i. 51. 



f Home, as cited in Griffith's 'Animal Kingdom,' vii. 98. 



I ' Bataviaasch Genootschap,' iii. 97. 



§ J. A. S. XV. 363. 



