1 30 Summer Studies of Birds and Books CHAP. 



first task will be to fix, by means of a delicate pitch- 

 pipe, the note on which the bird starts his song, and 

 so, as far as may be possible, to fix the key or 

 tonality of it. But this, though only the initial step, 

 is often extremely difficult to accomplish ; and you 

 can by no means always be sure that, even if there be 

 a real tonality in the strain, it can be determined by 

 making sure of the first note. And when you go 

 further and try to photograph on paper the whole 

 strain or even a part of it, you are confronted with 

 such difficulties as can be overcome by nothing 

 less powerful than a lively imagination. My own 

 attempts of this kind have been signal failures ; but 

 I should not be justified on that account in thinking 

 the task a hopeless one, if I were not wholly 

 unable to appreciate the attempts which others 

 have thought to be successful. To take an example : 

 I have seen the song of the Willow- wren represented 

 by a descending scale of notes from E in the treble 

 to C ten notes below, and I quite understand why 

 this succession of notes is selected. But I wholly 

 deny that the Willow-wren prefers to sing in C 

 major, and I much doubt whether there be in his 

 song one to which I must have listened many 

 thousands of times more than one full tone of our 

 musical scale. 1 This is one of those songs which 



1 The notation of this song will be found in Mr. Harting's Birds 

 of Middlesex, p. 58. In the same book is an attempt to reproduce 



