128 RELATION OF SONG TO THE TERRITORY 



answers the same purpose as the topmost branch 

 of a tree. 



Now there is nothing in the external 

 environment to which the song is more 

 definitely related than to the " headquarters "- 

 this at least is the conclusion to be drawn from 

 the behaviour, and I will indicate the sort of 

 evidence upon which such conclusion is based. 

 There is, first of all, the persistency with which 

 the 4 male resorts to the same tree, even to the 

 same branch, and, as it seems, solely for the 

 purpose of advertisement. We know by 

 experience the approximate routine of the 

 male's behaviour ; we know where to seek it, 

 where to hear it, and when once we have 

 discovered its headquarters, we know that there 

 it will sing day after day for weeks or it may 

 be fo* months together perhaps the most 

 striking feature of its behaviour at this season. 

 Next, we find that other trees, though made use 

 of, are not made use of to a similar extent for 

 the purpose of song. The area occupied varies 

 much according to the nature of the environ- 

 ment ; it is sometimes extensive, and seldom less 

 than half an acre or so in extent ; but in most 

 instances it contains plenty of trees and bushes 

 which could, one would imagine, serve the 

 purpose of a "headquarters" just as well as the 

 particular one selected, and yet the bird, when 

 there, betrays no inclination to sing at all com- 

 parable with that which can be observed when 

 it occupies its accustomed perch. Further 

 evidence is afforded in the behaviour of those 



