I 9 4 EVOLUTION OF BIRD-SONG 



species ; for this reason I have adapted a name for 

 the cry, and have called it " warblers' tewy" An 

 imitative bird will often mimic two or more cries 

 of others in one phrase : I have taken no regular 

 account of this, nor of the number of phrases 

 betraying no mimicry. The latter are heard in 

 about the ratio of one to three in phrases of the 

 thrush and robin, and nearly one in two of those 

 of the redstart, while there are practically none in 

 those of the starling and sedge -warbler. The 

 degree of practice performed and this is greatly 

 dependent upon the state of the weather and the 

 abundance of food has much influence on the 

 fluency of song and mimicry : after prolonged and 

 severe frosts I have heard most feeble and limited 

 songs, as well from starling as from robin and 

 thrush. 



THE THRUSH 



During fourteen months I listened to an aggre- 

 gate number, approximately, of 50 thrushes, and 

 I heard 1120 phrases * which seemingly contained 

 some recognisable imitation, and 450 in which I 

 detected none. The six subjects most frequently 



1 As already stated, the term "phrase" does not refer to any 

 particular strain or order of cries. A bird might repeat the same 

 strain many times, with alternate pauses, and each utterance would 

 be considered a phrase, and recorded as such. 



