148 U.S. NATIONAL MUSEUM BULLETIN 23 7 i'akt i 



not necessarily high, perch from which to sing, and I heard one bird 

 sing from the ground. In general, the song shows a tendancy to 

 start on a higher pitch and end on a lower one, seven of my records 

 ending on the lowest note of the song." 



Food. — ^A member of the Fringillidae, the nonpareil is primarily a 

 seed-eater. This has been shown by stomach analyses, although 

 these have not been extensive, and by observation. 



In South Carolina, I have observed the bhd closely. Its frequency, 

 not to say constant presence, in weedy fields, edges of woods and 

 salt marshes, roadside hedges, and so forth, indicate that seeds of 

 grasses are its main dependence. And added to the habitat is the 

 actual sight of the bird on the stems of such growth, picking away 

 at the heads of seed. Foxtail grass (Alopecurus) , some pines, figs, 

 and sunflowers furnish most of the preferred seeds in the Carolina 

 Low Country. 



In Florida, Howell (1932) examined 13 stomachs, in which vegetable 

 food composed 73 percent of the total content. Among the "con- 

 siderable quantities" of seeds were those of various grasses, sedges, 

 and weeds, including dock (Rumex acetosella), Panicums, Hypericums, 

 and Cyperus. One stomach contained pine seeds, another rose seeds. 

 Wheat was found in two, and fig seeds and pulp had been taken "in 

 several instances." Animal matter consisted of insects, amounting 

 to 27 percent; those represented were "beetles, grasshoppers, crickets, 

 bugs, wasps, flies and lepidopterus larvae." 



W. L. McAtee (in Beal, McAtee, and Kalmbach, 1916) says: "Few 

 complaints have been lodged against the painted bunting on the score 

 of its food habits. It is said to eat rice at times, to peck into figs and 

 grapes, and to bite off the tips of pecan shoots. In no case that has 

 come to notice, however, has it been charged \vith doing serious 

 damage. Certainly no such charge is supported by the investigations 

 of the Biological Survej^, for no product of husbandry has thus far 

 been found in any of the stomachs examined, 80 of which have been 

 examined, all collected in Texas in July, August and September." 

 He goes on to say that animal matter composed 20.86 percent and 

 vegetable matter 79.14 percent, closely paralleling the ratio in the 

 specimens mentioned by Howell from Florida. With the eastern and 

 western portions of the range thus indicated, it is not to be supposed 

 that much variation takes place in the central areas of the bird's 

 occurrence. 



McAtee further observes that 2.48 percent of the animal food "was 

 made up of weevils, mostly cotton boll weevils. All insects of this 

 group are destructive, but none more so than the notorious cotton boH 

 weevil, and this species had been eaten by 18 of the 80 nonpareils 

 examined." The cotton-worm is also eaten, and composed 3.14 per- 



