356 HIRUNDINTD/E. 



after it has reached us it commonly keeps pretty close to 

 streams or ponds, over which it may be seen taking its food 

 with the curious jerking flight, so well noted by Gilbert 

 White, and it does not seek its breeding-quarters till towards 

 the end of April or beginning of May. 



Like the species already described, this bird comes 

 to us no doubt from Africa, and almost always chooses its 

 nesting-place in the banks of rivers, sand-pits, railway- 

 cuttings, and other vertical surfaces of earth of a nature 

 that will enable it to perforate them for its purpose. In 

 such situations it bores horizontal galleries with a degree of 

 regularity, and an amount of labour, rarely exceeded among 

 birds. The mode in which these holes are made has 

 been described more or less fully by White, Rennie, and 

 Macgillivray's correspondents — Messrs. Duncan and Weir. 

 When beginning its excavation, the bird clings to the face 

 of the bank, steadying itself by its tail, and, using its bill 

 as a pickaxe, loosens the earth, which at first falls down by 

 its own weight clear of the hole. In doing this the bird 

 works from the centre outwards, assuming all sorts of posi- 

 tions, and as often as not hangs head downwards while 

 grasping the circumference with its claws. When the hole 

 is carried further the same method is pursued, but the 

 detached soil has then to be scraped out by its feet, since 

 the gallery though generally slopiug upwards from the 

 entrance is too nearly horizontal for the earth to run out of 

 itself. The form of the boring and its length seem to 

 depend much on the nature of the soil. Dry, fi'iable sand, 

 though easily pierced, has its disadvantages in the crumbling 

 of the sides, especially as the bird is breaking ground, till a 

 large irregular hole is made, and then the burrow is extended 

 perhaps to four, six, or, as one authority says, even nine feet. 

 Harder sand, lying often in layers, produces shorter tunnels, 

 from eighteen inches to three feet in length, with an oval or 

 oblong section, and it is only in very tenacious soil that the 



graphs are nothing bat Sand-Martius, the difference in the plumage and flight of 

 the two species, obvions at a glance to an expert, lieing unknown to the casual 

 observer. 



