Impressions of the Voices of Tropical Birds 



165 



were neither mills, steam nor villages in these mountains, which are un- 

 broken virgin forest for a hundred miles or more either way. Perhaps I had 

 heard a cicada. I could scarcely credit a bird with such a prolonged sound 

 as this. 



The next day I went back to solve the thing. When, after two hours 

 of steep ascent, I had reached the 8,000-foot level, I heard again my mysterious 



ANT-THRUSH {Formicarius rufipectus carrikeri) 



whistle. Listening carefully, and imitating it as well as I could, I was able to 

 discern that the sound became definitely more loud and distinct. No insect, 

 this. Soon I could analyze it quite closely, and found it to be a very gradually 

 rising crescendo, beginning about on C, and a full though slightly throbbing 

 or tremolo whistle. I was astonished at its duration, for I could detect no time 

 at which a breath could be taken. Timing three successive songs, I found 

 them to endure forty-seven, fifty-seven, and fifty-three seconds! This was 



