THE SECRET OF THE WH.LOW WREN 123 



no regular succession of intervals. Speaking of 

 the willow wren in this connection, he adds : 

 " Strange as it may seem, the songs of birds may 

 perhaps be more justly compared with the human 

 voice when speaking, than with a musical instru- 

 ment, or with the human voice when singing." 

 The truth of this observation must strike any 

 person who will pay close attention to the singing 

 of birds ; but there are two criticisms to be made 

 on it. One is that the resemblance of a bird's 

 song to a human voice when speaking is confined 

 to some or to a few species ; the second is that 

 it is a mistake to think, as Mr Fowler appears to 

 do, that the resemblance is wholly or mainly due 

 to the fact that the bird's voice is free when sing- 

 ing — ^that, like the human voice in talking, it is 

 not tied to tones and semitones. For instance, 

 we note this pecuHarity in the willow wren, but 

 not in, say, the wren and chaffinch, although the 

 songs of these two are just as free, just as inde- 

 pendent of regular intervals as our voices when 

 speaking and laughing. The resemblance in a 

 bird's song to human speech is entirely due to the 

 human-like quality in the voice ; for we find that 

 other songsters — notably the swallow — have a 

 charm similar to that of the willow wren, although 



