IMITATION OF BIRDS 193 



can only be detected by a listener who has an 

 accurate memory for the sounds mimicked, and a 

 possesser of this faculty will the more easily per- 

 ceive imperfect imitations in proportion to its 

 power. 



But it is impossible for the listener educated in 

 these matters to teach his knowledge to other people : 

 it can only be gained by close personal attention to 

 the subject, by identifying birds when they exclaim 

 or sing, and by a sufficient power of memory. If, 

 therefore, a person who has studied the voices of 

 birds should claim to have heard certain imitations, 

 the reader must form his own estimate of the state- 

 ments, which cannot, except in very rare instances, 

 be subjected to proof; but when several well-known 

 authorities have stated (as already quoted) that they 

 had observed similar incidents, the statements must 

 be considered as so much the more acceptable. 



It should be observed in this place that some 

 seeming imitations are difficult to relate to one 

 particular bird, as they are like a cry common to 

 two or more species. Thus, as has been noticed, 

 the little upwardly -slurred whistle which is the 

 alarm-cry of both chiffchaff and willow -warbler, 

 and is uttered more loudly by the nightingale, 

 cannot, when imitated, be related to either of these 

 13 



