NESTING OF THE ROUGH-WINGED SWALLOW 443 



intervals along the banks of streams, wherever favorable sites 

 occur. The birds seldom excavate holes for themselves, pre- 

 ferring to take some suitable cavity and refit it to their taste; 

 thus, they are often found in deserted Kingfishers' holes, where 

 the nest is placed a foot or so from the entrance. They will 

 also, on finding a decayed root of sufficient size leading in from 

 their favorite sand banks, remove the soft puuky wood, follow- 

 ing the winding of the root to a depth of about two feet, where 

 they place the nest in an enlarged cavity. Besides this, they 

 like to build in holes in masonry, near water. In the few 

 observed instances of their digging a hole for themselves, they 

 worked in rather a slovenly way, making holes larger than 

 appeared necessary, and invariably circular at the entrance — 

 the Bank Swallows' holes, on the contrary, being quite sym- 

 metrically elliptical, with the longer axis horizontal, and no 

 larger than required for the free i)assage of the birds — too small 

 to admit the hand, while the Eough-wings' nests may usually 

 be reached without difiiculty, except when built in masonry, in 

 which latter case the birds may pass through a crevice barely 

 wide enough to admit them, providing the cavity within be 

 suitable for a nest. The nests of serripeunis are more carelessly 

 constructed, as a rule, than those of riparia are ; the birds do 

 not seem to search at any distance for particular materials, 

 being satisfied with anything that may be at hand. One nest 

 built in a Kingfisher's hole in a sand bank about fifteen rods 

 from a poultry yard, was composed entirely of the feathers of 

 domestic fowl. In another instance, three fresh eggs were 

 found on the bare sand, in a mere pocket barely six inches 

 deep, indicating that the mother bird was so pressed to lay 

 that she had no time to complete her nest. Not infrequently 

 fresh eggs are found in the same nest with others far advanced 

 in incubation, and occasionally fresh eggs, others newly 

 hatched, and young birds may be found together. 



Other writers witness a still wider range of variation in the 

 nidification of the Rough-wings. Cooper speaks of their nest- 

 ing in California in burrows in sandy banks, two or three feet 

 deep, closely crowded together, and near the upper edge of 

 the embankment; as well as of their resorting sometimes to 

 natural clefts in banks, in adobe buildings, and even in knot- 

 holes. Their breeding in the last-named places is probably 

 exceptional, but it is known that even the Bank Swallow, the 

 most inveterate and conservative of the family, will sometimes 

 take to a tree, and Denshaw furnishes probable confirmation 



