Birds of Gazaland. 103 



in search of nesting-sites. In November breeding is in full 

 swing in such places. The nest consists of a broad chamber 

 of mud pellets thickly lined with soft grasses and other 

 material, and is entered from the side through a tunnel of 

 variable length, sometimes a foot or more. Occasionally 

 it is attractively constructed of clay of as many as three 

 different colours. On November 7th, 1906, I had just at 

 dusk shot a Terpsiphone, with a view to identifying its nest, 

 wheu I noted that its mate, which was continuing to call, 

 was being answered each time from a small cliff close by 

 by the somewhat nasal "weeping" note of a Swallow, and 

 1 found on investigating that a pair of Swallows of the 

 present species were resting side by side in tl-.e chamber of 

 their nest under a ledge of rock. There were no eggs. 



These Swallows are always attracted by grass-fires, and the 

 following extract from my diary for September 5tli is illustra- 

 tive of their usual behaviour on such occasions: — '^Wheu we 

 commenced to burn, first two Swallows appeared, then a few 

 more pairs, and soon an enormous flock was present — not 

 less, I should say, than two hundred, — which, keeping more 

 or less together, hawked up and down over and around the 

 smoke, never venturing into the denser cloud and seeming 

 to get plenty of insects. After a time they became slacker 

 in their hunt and took to flying higher, amusing themselves 

 by fluttering up against the wind time after time in a fairly 

 compact body, being evidently by this time replete. Finally, 

 they rose and made off, the last I saw of them being at some 

 distance from the fire, wheeling idly about in corapauy with 

 a Kite at a great height in the air. They were practically 

 all H. puelia, but I once or twice saw a small Swallow which 

 I took to be H. dimidiata." The stomach of a female shot 

 on this occasion contained two Coccinellidse, a small green- 

 winged neuropterous insect, and a frog-hopper. A few^ pairs 

 of Hirundo puelia also come to the smoke whenever I burn 

 my seed-beds, but they at once make off on finding that no 

 insects fly out ; and this action on their part must com- 

 municate the fact to the others, for no matter how much 

 longer the fire continues no further Swallows visit it. 



