256 THE BIRDS OF THE BRITISH ISLES. 



many of these tunnels are visible in the steep cuttings of the 

 Manchester Ship Canal. Perhaps the birds do not habitually 

 roost in the tunnels during construction, but I have taken them 

 at night from incompleted burrows. In the fatally late spring 

 of 1886, when a hopeful advent was followed by a severe snow- 

 storm in May, large numbers took refuge in the burrows and 

 perished ; Mr. Oldham took fifteen bodies from one hole. Both 

 birds excavate, boring with the beak and scratching out the 

 loose sand with their small feet. The actual nest is a litter of 

 straw and feathers in a chamber at the end of the burrow ; it is 

 soon a hotbed of parasites. On the coast I have found nests 

 entirely composed of seaweed and gulls' feathers. Four or five 

 white eggs (Plate 84) are laid about the middle of May, and 

 a second brood is usual. The colony ^Plate 105) is a busy 

 place ; the young, showing their white chins, peep out of the 

 mouth of the hole and clamorously demand food. At intervals 

 the birds go through a strange performance, common to other 

 colonial nesters such as gulls and terns. The air will be full of 

 bustling, twittering birds when suddenly there is a strange hush, 

 and all, as if by word of command, rise high into the air and 

 leave the place deserted. After a short interval they return a 

 few at a time and activities are resumed. This may bo aerial 

 play, but the silence and sudden departure giv^e the impression 

 of alarm, though no cause is apparent ; it almost suggests 

 emergency practice, like boat- or fire-drill. The Sand-Martin 

 is victimised by the House-Sparrow which sometimes occupies 

 holes in the midst of a colony, and I have seen one steep bank 

 w^here every burrow was stuffed with the Sparrow's untidy litter 

 and all the rightful owners evicted. 



The Sand-Martin departs early, at any rate from its more 

 northerly haunts. In August the gatherings at the nightly 

 roost increase enormously, though the advent and departure of 

 passage birds causes great irregularity in numbers. It is then 

 that the crepuscular flights become most interesting. From 



