THE SWALLOWS 261 



broods and have the whole process completed 

 before August begins, while it is no uncommon 

 thing for swallows to be hard at work right up 

 to the end of September. Sometimes, indeed, the 

 second brood is unfledged when the time for the 

 migration comes round, and in such a case as 

 that the old birds have been known to go with 

 their kind, leaving the young to die of starvation in 

 the nest. The fact gives one a vivid sense of 

 the impelling character of the instinct which can 

 overcome even that of parental affection, so very 

 strong in all birds ; but the prolongation of the 

 nesting season remains unexplained. 



It has been suggested that the explanation may 

 be found in the fact that the nest -making process 

 in all three species of swallow is a very prolonged 

 and elaborate affair. This is true. The sand- 

 martin, the weakest of them all, drives a tunnel 

 often four or five feet long into banks of sand 

 sometimes approaching rock in its compactness. 

 When it is remembered that all this is accomplished 

 with the soft and short beak of an insect-eating 

 bird as the only tool, it is possible to realize the 

 magnitude of the labour. Nor is the driving of 

 the tunnel the whole work, for at the end of it 

 the martin builds a nest which, though not an 

 elaborate piece of architecture, contains a large 

 mass of fibrous and feathery material which takes 

 time to collect. 



The nests of the swallow and the house-martin 

 are of a totally different design from that of the 

 sand-martin, but each in its building takes up 

 much time. Both the saucer-shaped nest of the 



