552 JOURNAL, BOMBAY NATURAL HISTORY SOCIETY, Vol. X 



walls only at the very edge alone being of the usual thin character* In 

 its greater external length the nest was so less than two inches, and it 

 stood out from the tree rather less than an inch and-a-half. Externally 

 it was nearly two inches in depth, the nest running down the tree in 

 a kind of irregular cone with the base reversed. Inside, it was less 

 than half an inch in depth. 



They breed principally in December and January. It is a most 

 interesting sight to watch them retiring to roost. Just after sunset 

 one hears a harsh scream overhead, then another, and another, as one 

 by one the birds collect. On arriving at their proposed roosting place, 

 they fly round and round, gradually lowering their flight until one 

 bird suddenly makes a swoop and settles on some part of the tree 

 near the top. This is the signal for the rest to perch, and in a few 

 minutes they are all dotted about the higher branches. In a few 

 minutes they begin to close up with the bird which first alighted on 

 the tree, finally collecting in a feathery ball, one on the top of the other. 

 The first attempt is, however, very seldom satisfactory, and abruptly 

 scattering the whole performance is again gone through ; sometimes 

 this occurs three or four times before they get settled, but at last the 

 twittering stops, and they are asleep for the night. It is wonderful 

 how compactly these birds close up ; a flock of eleven appeared not 

 to take up a space more than a foot long by about half as broad. 



Anyone who knows Artamus fuscus (or any other Swallow-shrike) 

 well will notice at once that the above notes on the roosting of the 

 Swifts would equally apply to that bird ; there is, however, a very great 

 difference in the speed at which the two birds fly ; and again whereas 

 Artamus fuscus collects in flocks, sometimes numbering over a hundred, 

 those of M. covonata will seldom be over a dozen or so. I have spoken 

 above of a flock of eleven, and of these six were shot, five of the birds 

 provino- to be males and one an immature female. As they vi^ere shot 

 in January, this would seem to prove that the female must have been 

 a bird of the season hatched elsewhere, and that this Swift has two 

 broods and visits two breeding places in the year. It would appear 

 that the adult breeding males collect together to roost, leaving theiT 

 better halves to perform the duties of incubation during the night. 



Hume (" Stray Feathers," Vol. XI, page 33), remarks that it is 

 strange that this bird should not be foimd in any of the hills round 



