184 BULLETIN 196, UNITED STATES NATIONAL MUSEUM 



birds to be seen are generally not singing, and the song is not heard till 

 the species has been present for several dates. The average date for 

 the first bird seen, in my records, is May 13, but the average date for 

 the first song is May 19. Generally the last birds to be seen, late in 

 May, are still in song. 



"On the breeding grounds the birds sing till late in July or early in 

 August; the average date of the last song in 15 years of observation is 

 July 30, the earliest July 12, 1929, and the latest August 8, 1928. At 

 Flathead Lake, Mont., where the oliveback is the most abundant 

 breeding bird, 12 different individuals could be heard in the evening 

 from the Biological Station. They sing later in the evening than any 

 other diurnal bird. In Allegany State Park, N. Y., where four species 

 of thrushes can be heard in the evening chorus, the last one to sing, as 

 darkness comes on, is the olive-backed thrush. In the Adirondacks I 

 met a party of people who mistook its late singing for that of the whip- 

 poorwill." 



As an aid in remembering the song, A. D. DuBois (MS.) thinks of it 

 as saying "whip-poor-vrill-a-will-e-zee-zee-zee, going up high and fine at 

 the close. Sometimes there is an extra a-will." Albert R. Brand 

 (1938), who has recorded on films the virbration frequencies in the 

 songs of so many birds, found for the olive-backed thrush an approxi- 

 mate mean of 2,925, the highest note 3,850 and the lowest 2,000 vibra- 

 tions per second; the high note is considerably lower than that of either 

 the hermit or the veery, but the low note of the hermit is much lower 

 in pitch than either of the others. Comparison with the high frequen- 

 cies of two of the shrillest singers is very striking. The corresponding 

 figures for the grasshopper sparrow are 8,600 mean, 9,500 high, and 

 7,675 low; and for the blackpoll warbler the mean is 8,900, the high 

 10,225, and the low 8,050 vibrations per second. 



Stewart Edward White (1893) studied the song of one of these 

 thrushes that sang regularly near his cottage in Michigan; he figured 

 that this bird sang 4,360 songs per day, and writes: "He sings on an 

 average nine and a half times a minute with extreme regularity. 

 During the song periods of morning and evening his constancy of 

 purpose is remarkable; except to seize a passing insect, he never 

 breaks the regular recurrence of his song. From a series of records 

 it is found that he begins on an average about 3.15 a. m., and sings 

 steadily (of course by that I mean ten times a minute, not con- 

 stantly) until about 9.00 a. m.; he is nearly silent until noon, after 

 which he sings occasionally for a minute or so. About 4.30 he begins 

 again, and only ceases to retire for the night about 7.30 p. m." 



Only a few observers have written about the night singing, but 

 Dr. Wilfred H. Osgood (1909) has given us a pleasing account of 

 hearing such a concert in Alaska: 



