394 THE ARCHITECTURE OF SWALLOWS 



faces, whether artificial or natural, aud loosely furnished with 

 soft material. This is seeu iu perfection in the nesting of 

 Petrochelidon lunifrons, and is imperfectly illustrated by the 

 nidification of Hirundo horreorum. It has every appearance of 

 an acquired trick, by which these hole-breeders carry out their 

 original instincts on a greatly improved plan. 



It should be observed that all of our Swallows have been 

 modified by human agency excepting the Bank Swallow, nest- 

 ing diflereutly at successive periods; that some of them, like 

 the Purple Martin and the Violet-green Swallow, are still 

 serving their apprenticeship under the new regime which the 

 settlement of the country has brought about; and that even 

 the Barn Swallow, which seems so thoroughly changed, still 

 sometimes nests in the West in its primitive fashion. Those 

 whose acquired habits have become thoroughly ingrained are 

 now pretty constant in their adherence to a single plan of 

 architecture; but the Violet-green Swallow, for instance, at 

 present nests in a very loose fashion, according to circum- 

 stances. 



There is no question of the fact, that some of the Swallows 

 which in the East now invariably avail themselves of the 

 accommodations man furnishes, in the West live still in holes 

 in trees, rocks, or the ground. Thus I have found Purple 

 Martins breeding in Arizona nowhere but in the hollows of 

 trees, chiefly deserted Woodpeckers' holes ; and I have known 

 Barn Swallows to nest on embankments of earth. In the 

 West, again, where the Cliff Swallows habitually affix their 

 nests to the irregular surfaces of clilfs, retort-shaped structures 

 are not so often perfected as when the birds build against the 

 smooth sides of houses. In this case, however, it may be a 

 question which is the original method, which the later modifi- 

 cation. For in many special cases, the Eave Swallows have 

 been known to build mere open cup shaped nests affixed to 

 buildings, quite like the rafter-nests of the Barn Swallow, 

 when they had similar bases of support to work upon. 



There is another curious fact, which shows that the nidifica- 

 tion of Swallows may be changed and improved, without any 

 variation in the character of the site selected, but apparently 

 as the result of reflection and some degree of ratiocination. 

 This is the case with the common House Martin, Chelidon urhica. 

 I quote the words of Dr. Brewer (Am. Nat. xii. 1878, 30) : — " A 

 few years ago it was discovered by accident that within fifty 



