CHARACTERS OF THE GENUS PETROCHELIDON 425 



eggs, as bulky structures of mud, aud like those of Cliff Swal- 

 lows. Being perfectly familiar with the birds, he could uot 

 have beeu mistaken in identifying the species ; and he agreed 

 ■with me that the birds must have occupied in these instances 

 the deserted nests of other Swallows. This brings up Nuttall's 

 early testimony to the same effect, and makes it seem much 

 more probable — il" it may not indeed be regarded as confirma- 

 tory — though he or Townsend certainly got hold of the wrong 

 egg, a drawing of which subsequently came into Dr. Brewer's 

 possession through Audubon. We should expect the Violet- 

 greens, on yielding to civilization, to come to terms in the 

 same way the Martins and White-bellies have, by occupying 

 boxes set up for their use, or else to enter knot-holes or the 

 crevices behind weatherboards, as the Wrens; but that their 

 habits will be modified in some way, and at no distant day, 

 there is no reasonable doubt. With which understanding, I 

 leave the wilful and capricious little creatures to enjoy their 

 hermitages, whether of tree or rock, as long as they please. 



Genus PETROCHELIDON Cabanis 



Hirundo, p., of Authors. 



Petrochclidon, Cab. MH. i. 1850, 47 (type H. melanogaster Sw. = P. swainsoni Scl.). 



This is one of the better-marked groups which have been 

 established as generically distinct from Hirundo. In some re- 

 spects, it rather approaches Progne. The bill is quite stout and 

 deep (for this family), and the nostrils are superior, opening 

 without nasal scale. The tail is unusually short, the tips of 

 the folded wings reaching beyond it, and is about even, or only 

 slightly emarginate, with the feathers broad to their ends. 

 The feet are much as in Hirundo ; the tarsi are feathered above, 

 and the toes are extensively adherent at base. There is a 

 certain bristly appearance of the front and chin, different from 

 what is seen in other groups. The tuft of crissal feathers is 

 full, reaching nearly to the end of the tail. The species agree 

 well in a special pattern of coloration, being steel-blue above, 

 with rufous rump and nuchal band, and usually a frontlet of 

 different color from the rest of the upper parts ; the under 

 parts are not continuously white, as in Tachycineta. The nidi- 

 fication of some, if not all, of the species, is peculiar. The eggs 

 are colored, as in Hirundo. 



The species are characteristically American, though one has 

 been described from the Cape of Good Hope. An Australian 

 species referred to Petrochelidon by ('abanis is by Gray ranged 



