136 Summer Studies of Birds and Books CHAP. 



It was based chiefly on observations made upon birds 

 in captivity a kind of evidence constantly apt to be 

 misleading. It is, in fact, extremely difficult to ob- 

 tain really trustworthy observations as to the court- 

 ing of birds in a wild state ; even if we were to spend 

 whole days in watching them from sunrise to sunset, 

 their restlessness and locomotive power would make 

 it very hard to prove beyond doubt that a given 

 female selects the best singer among her wooers, nor 

 could we feel very confident that the best singer in 

 our judgment is also the best in hers. And Darwin, 

 I think, was well aware of the doubtful character of 

 his evidence; he records what bears against his 

 theory with all his usual candour, and does not 

 really press this part of his argument home. There 

 are, however, two unquestionable facts which sustain 

 his view. In the first place, true song is always a 

 male character ; if female birds sing at all, they sing, 

 so far as seems to be known, a feebler and inferior 

 song, which may be no more than a humble imitation 

 of that of their lords. And secondly, the song of 

 the male is at its best during the breeding season, 

 and in many or most cases is heard at no other time. 

 To these two facts we may add another, less cogent 

 but very interesting, that most of our best singers 

 are very quietly-coloured birds ; the inference here 

 being that where beauty of plumage has not been 

 developed to charm the taste of the females, another 



