24 CTPSELIDJE. 



and houses, in the corners of old stone buildings, and in verandahs, 

 either inside or outside, if there is protection from sun and rain. 

 Various observers describe the nest as somevrhat differently con- 

 structed. Burgess says that he has seen their nests crowded 

 together under the roofs of old buildings, choultries, and temples ; 

 one nest from a rock was built of mud, hned with grass, and con- 

 tained two white eggs. Layard states that in Ceylon they breed in 

 great numbers on rocks, also under bridges, and that the nests 

 built in clusters are composed of mud and grasses, with a small 

 round entrance, precisely resembling those of the Martin (//. 

 rirhica) ; the eggs, from two to four in number, pure white. Adams 

 says that the nest is of mud, mixed with wool and feathers. In 

 some of these cases the great A\eight and solidity of the nests may 

 have led the observer to conclude that they were made of mud. 

 The nest has generally a slight hollow in one place for the recep- 

 tion of the eggs, which are usually two in number, sometimes three, 

 and pure white." 



1 may here add that I have seen not hundreds, but tens of 

 thousa,nds of these nests in all parts of Continental India ; and 

 that, like Jerdon, I never knew of this species using any mud in 

 the construction of its nests. 



Mr. Gt. Vidal writes from the South Konkan : — " Common 

 throughout the seaboard. Nests found in Eebruary and April in 

 clusters on the island fort of Suvamdurg and the rocky cliffs on 

 the coast." 



Mr. "W. V. Legge, writing from Ceylon, says of the breeding of this 

 Swift : — " The Indian Swift breeds in February and March in the 

 south-east of Ceylon, nesting under bridges and in the roofs of 

 outhouses. I found a large colony in the month of March, 1872, 

 nesting under the tiles and between the rafters of the roof of the 

 salt store at Hambautota. The nests were placed close together in 

 some instances, and \vere of all shapes and sizes ; they were con- 

 structed of grass and native cotton, and lined with feathers mixed 

 Mth the latter material. The eggs, in most instances, were three 

 in number." 



And Colonel Legge gives March to July as the limits of the 

 breeding-season in Ceylon. 



I think tliree is the normal number ; I have certainly far more 

 often found three, or even two, than four eggs. 



Typically, these eggs are excessively long and narrow o\als, 

 pointed towards one end, and often somewhat pyriform in shape. 

 They vary, however, much both in size and shape ; and of speci- 

 mens of my own taking, some are fully one third longer than 

 others, while the cubic contents of one egg I have must be fully 

 twice that of another. In colour they are of course a perfectly 

 pure and spotless white, with commonly scarcely a trace of gloss, 

 though occasionally a slightly glossy egg is met ^\'ith. Several 

 specimens I ha'\-e are fully as long as the egg of Gypselus apus 

 figured by Mr. Hewitson, but their breadth is rarely even five 

 sevenths of the one there represented. 



