208 BULLETIN 19 6, UNITED STATES NATIONAL MUSEUM 



Chrysomelidae, 18.6 percent; Lepidoptera larvae, 8.6 percent; Elater- 

 idae, 7.6 percent; Cerambycidae, 7.0 percent; Tipulidae, 6.4 percent; 

 Carabidae, 3.0 percent; Hymenoptera, except ants, 2.4 percent; Hom- 

 orus undulatus, 2.0 percent; Staphylinidae, 1.4 percent; Ephemeridae, 

 1.4 percent; miscellaneous Coleoptera, 1.0 percent; weevils, 0.6 per- 

 cent; Lampyridae, 0.2 percent; and Diptera, 0.2 percent. 



Their proclivity for ants is clearly shown, for in four of the birds 

 42 to 55 percent of the food consisted of these insects; the food of the 

 other bird, however, consisted of only 5 percent Formicidae and 90 

 percent Chrysomelidae, with the other 5 percent Elateridae. 



The list of food items found in two stomachs, taken on Mount 

 Mansfield in June and examined by Dr. Clarence Cottam, is given 

 in too much detail to be included here (see Wallace, 1939). The 

 items that occurred to the extent of 5 percent or more are as follows : 

 Carabidae, 24 percent, in one stomach only; Rhyacophilidae and 

 Trichoptera, 10 percent each; Calliphoridae and Anthomyiidae, 8 per- 

 cent each; Neuroptera and Phalangidae, 7 percent each; Diptera and 

 Syrphidae, 6 percent each; and Chrysomelidae and Empididae, 5 

 percent each. Formicidae amounted to only 1 percent and that in 

 only one stomach. 



Late in summer and in fall these thrushes become partly frugivorous, 

 taking considerable quantities of berries. Brewster (1906) writes that 

 in autumn they subsist largely on berries — cornels, deadly nightshade, 

 barberry, spicebush, wild grape, woodbine, and poison-ivy. Burns 

 (1919) reports collecting a specimen of bicknelli at Berwyn, Pa., on 

 October 6, "while feeding on poke berries." 



Nevertheless their predilection for animal food is probably never 

 lost, even in autumn, and what fruit they take seems more or less 

 incidental to their habitual animal diet. Experience with two caged 

 birds, though not an accurate criterion for determining what thrushes 

 eat in nature, at least shows certain food preferences. These birds ate 

 considerable quantities of the berries indigenous to Mount Mans- 

 field — blueberries, bunchberries, snowberries, red-berried alder, and 

 twisted-stalk — but they never could be induced to subsist entirely, or 

 even largely, upon them, even for a short time. With their cage well 

 supplied with berries they would clamor for insects and eat them 

 greedily when brought. During the fall and winter the surviving bird 

 lived chiefly on animal food (fresh meat) varied now and then with 

 fruit. This presumably indicates that berries and fruit, which really 

 may be consumed in surprising quantities at times, play only a 

 secondary role in the Bicknell's diet, serving merely to supplement 

 items of an animal nature. 



An unexpected taste, discovered as the result of caring for caged 

 birds, is an apparently natural desire for something leafy, two young 



