fwallow, was in his contemplation, becaufe he firfi fpeaks of the 

 bird's building ag.dnft windows, before he mentions chimnies, 

 and therefore fuppofes that either place is indifferent ; which is 

 not the cafe, becaufe the fwallow feldom builds on the fides of 

 windows, or the martin in chimnies. 



There are perhaps three or four martins toone fwallow in all 

 parts ; and from their being the more common bird of the two, as 

 well as from the circumflance of their building at the corner of 

 windows (and confequently being eternally in our fight) nine- 

 teen out of twenty, when they fpeak of a fwallow, really mean 

 a martin n . 



I only take notice of this fuppofed inaccuracy in Monf. de Buf- 

 fon, becaufe, if that able naturalifr does not fpeak of the different 

 forts of fwallows with that precifion which is neceffary upon fuch 

 an occafion, why mould he rely fo entirely upon the impoffibility 

 of M. Adanfon's being miftaken ? 



I mall now ftate the experiment of Monf. de Buffon, to prove 

 that the fwallow is not torpid in the winter, and muft therefore 

 migrate to the coaft of Senegal °. 



n In the fame manner the generical name in other languages, for this 

 tribe of birds, always means the martin, and not the iWaliow. 



Thus Anacreon complains of the %$X$cm for waking him by its twit- 

 tering. 



Now if it be conlidered that there was ofily the kitchen chimney in a 

 Grecian houfe, it muft have been the martin which built under the eves 

 of the bed-chamber window, that was troublefome to Anacreon, and 

 not the fwallow. 4 



Ovid alfo fpeaking of the neft of the hirundo, fays, 



— luteum fub trabe figit opus, 

 by which he neceffarily alludes to the martin, and not the fwallow. 

 Garrula qu^ tignis nidum fufpendit hirundo. 



Virg. Georg. l.IV. 



? Plan de l'ouvrage, p. 15. 



