BUZZARD'S REST. 53 



They were not disturbed, but left the neighborhood, 

 after so much labor, voluntarily. I mention this in- 

 stance, and it is not an unusual one, to show that for 

 weeks before they finally settle down to rear a family 

 these birds are mated, and if for weeks prior to nidifi- 

 cation, why not for a much longer period ? 



I am quite sure tliat during the winter they were 

 seen in pairs as frequently as singly ; and their predi- 

 lection for hollow trees, at all times, suggests that like 

 2)igeon-woodpeckers, they often prepare nesting places 

 which, before completion or directly after, are perma- 

 nently abandoned. 



It is impossible to believe that such a sentiment as 

 Platonic friendship ever entered the mind of any sj^ecies 

 of bird. 



The Carolina wren affords a most striking instance 

 of birds remaining mated from one year to another. I 

 have had a pair of these birds under almost daily ob- 

 servation for two years, and it matters not what the 

 season may be, there was never more than a week passed 

 that the pair were not seen together. They commence 

 nesting in April and raise four broods. 



A pair of cardinal - grosbeaks were found nesting, 

 June, 1883, and the female was readily distinguished by 

 a peculiarity in the coloring of her wings. The pair 

 remained in the locality during the succeeding win- 

 ter, nested in the old site in 188-i, and in 1885 chose a 

 new position in a thicket of smilax a few yards distant. 

 This pair of redbirds were always associated during the 

 two winters that I had them under observation. 



Meadow-larks, if not in loose flocks, as though two or 



