ANIMAL BEHAVIOR — WELLS 421 



songs. To hear the brilliant little song of a chaffinch, sung from a 

 branch just outside one's window as one lies in bed on a sunny April 

 morning, is delightful. To hear it again a second time, a few seconds 

 later, is very well. But if the bird repeats the identical song phrase 

 over and over again, at extremely regular intervals of about 10 seconds, 

 as it often does at that season — and let us remember that 10 sec- 

 onds is just long enough for one's thoughts to get nicely settled else- 

 where — one begins to wish that it would stop. 



Such an experience impresses upon us that the singing behavior 

 of our common woodland and garden birds is rhythmically organized, 

 and we can distinguish two levels in this organization. 



First, is the pattern of the individual song phrase. This is highly 

 specific. The biological function of song is above all to warn off 

 other birds of the same species and sex, and to inform them that the 

 singer has taken possession of the surrounding territory, which he 

 is prepared to defend. Evidently, the phrase must be recognizable 

 and distinctive of the species. There is of course a certain amount of 

 variation from individual to individual and from moment to moment 

 in any one bird, and in some cases (such as the chaffinch) there are 

 well-marked local dialects. Nevertheless, the songs of any one species 

 have enough in common to be readily identifiable as such, and a bird 

 can be known by its song as surely as it can by its plumage. 



To some extent these song phrases are learned from other birds, to 

 some extent they are innate. The extent to which song has to be 

 learned probably varies from species to species, but the innate con- 

 tribution seems always to be substantial. 



Dr. W. H. Thorpe has described the songs of chaffinches that were 

 hand-reared in soundproof rooms at Cambridge and isolated from 

 any contact with experienced birds since the first few days of nestling 

 life. When such chaffinches grow up and sing, all the finer details of 

 the adult song are lacking ; there is, nevertheless, a rough framework 

 of definite structure on which they could be hung. In his own words : 



The experiments with the hand-reared birds suggest that there is an inborn 

 basis to the song but that it is extremely generalized. Innately these birds 

 seem able to produce a song of about the normal length, 2-3 seconds, and 

 showing a tendency to crescendo accompanied by a fairly steady fall in mean 

 pitch . . . there is a clear tendency in some birds to conclude the song with a 

 single, simple note of a higher pitch than the rest. 



Such a framework is, in itself, quite a complicated thing to inherit. 



According to Dr. Sauer, of Freiburg, whitethroats reared in isola- 

 tion from the egg produce the entire song when the right time comes, 

 down to the finest detail. 



The second level of organization is the distribution of the song 

 phrases in time. Many birds, when in singing mood, will shoot off 



