42 BULLETIN 126, UNITED STATES NATIONAL MUSEUM. 



on beechnuts and acorns or to pick up an occasional slug, snail, frog, 

 or lizard. In the vSouth they resort to the rice fields and savannas 

 in large numbers, feeding both by day and night if not disturbed; 

 where they are hunted persistently they become more nocturnal in 

 their feeding habits. 



Mr. W. L. McAtec (1918) has j)ublished an exhaustive re})ort, 

 based on the examination of 1,578 gizzards of the mallard by the 

 Biological Survey, from which I quote as follows: 



Approximately nine- tenths of the entire contents of tlie 1,578 mallard Bloraaclis ex- 

 amined was derived from the vegetable kingdom. The largest proportion of th-e food 

 drawn from any single family of plants came from the sedges and amounted to 21.62 

 per cent of the total. Grasses rank next in importance, supplying i:>.39 per cent; 

 then follow sraartweeds, 9.83; pond weeds, 8.23; duckweeds, 6.01; coontail, b.dl; 

 wild celery, and its allies, 4.26; water elm and hackberries, 4. 11; wapato and its 

 allies, 3.54; and acorns, 2.34 per cent. Numerous minor items make up the remain- 

 der. Some of the stomachs of the mallards were interesting on account of the large 

 numbers of individual objects they contained. For instance, one collected at Ham- 

 burg, La., in February, revealed about 28,760 seeds of a bullrush, 8,700 of another 

 sedge, 35,840 of primrose willow, and about 2,560 duckweeds as the principal items, 

 a total of more than 75,200. 



The animal food of the mallard duck though extremely varied may be classed in 

 five main groups: Insects, which constitute 2.67 per cent of the total diet; crus- 

 taceans, 0.35; molUisks, 5.73; fishes, 0.47; and miscellaneous, 0.25 per cent. 



Dr. Thomas S. Roberts (1919) has published the following inter- 

 esting note, showing the useful work done by mallards in destroying 

 mosquitoes. 



The late Dr. Samuel G. Dixon, while health commissioner of Pennsylvania, pub- 

 lished an article in the Journal of the American Medical Association for October 3, 

 1914, detailing results of experiments made by him along this line Two dams were 

 constructed on a stream so that the ponds would present exactly the same conditions. 

 One was stocked with gold fish and in the other 20 mallard ducks were allowed to 

 feed. After several mouths the duck pond was entirely free from mosquifoea while 

 the fish pond " was swarming with young insects in different cycles of life.' Ten 

 well-fed mallards were then admitted to the infested pond. At fiist they were 

 attracted by the tadpoles but "soon recognized the presence of larvae and pupae of 

 the mosquito and immediately turned their attention to these, ravenously devouring 

 them in preference to any other food present. At the end of 24 hours no pupae were 

 to be found and in 48 hours only a few small larvae survived. 



Mr. Edward H. Forbush (1909) says: 



It sometimes attacks sprouting or ripened grain but like most fresh-water fowl it 

 is undoubtedly of service in destroying such insects as tiie locusts and aiTny worms 

 which sometimes become serious pests. Professor A ughey found in the stomachs of 

 ten mallards taken in Nebraska 244 locusts aud 260 other insects, besides mollusks 

 and other aquatic food. 



Mr. J. H. Bowles (1908) records an interesting case of lead })oison- 

 ing among mallards which had been feeding in a marsh that for 

 many 3^ears had been a favorite shooting resort. The ground must 

 have been thoroughly sprinkled with shot for the stomachs of the 



