14 ON THE MIGRATION OF BIRDS. 
- 
nests in the same situations, this alone is a strong 
proof of their identity: great additional weight, how- 
ever, is given to this proof by the peculiarity of the 
situations in which such birds occasionally build. 
For three successive years a pair of Swallows: built in 
a pigsty belonging to a relation of mine, their ingress 
and egress being by a very low entrance; and in 
Bewick’s ‘ History of British Birds,’ vol. i. p. 253, it 
is stated, on the authority of Sir John Trevelyan, Bart., 
that “at Camerton Hall, near Bath, a pair of Swal- 
lows built their nest on the upper part of the frame 
of an old picture over the chimney, coming through a 
broken pane in the window of the room. They came 
three years successively, and, in all probability, would 
have continued to do so if the room had not been 
put into repair, which prevented their access to it.” 
White, in speaking of the Selborne Swifts (‘ Nat. 
Hist. Sel.’ p. 186), says, “they frequent in this vil- 
lage several abject cottages; yet a succession still 
haunts the same unlikely roofs: a good proof this,” 
he observes, “‘ that the same birds return to the same 
spots.” And he remarks of the House-Martin (p. 161), 
that “the birds that return yearly bear no manner 
of proportion to the birds that retire;’’ and this is 
uniformly the case. Now Swallows and House- 
Martins have frequently two broods in a summer, 
the first consisting of about five young ones, and the 
second of three, upon an average; and Redstarts, 
Spotted Flycatchers, and Swifts have one brood, the 
