140 BULLETIN 126, UNITED STATES NATIONAL MUSEUM. 



shoulders and wings; the fully adult dress is not acquired until the 

 following October or at the age of 17 months. Millais (1902) also 

 says that the immature females do not breed during their first spring. 



The midsummer eclipse plumage of the male is quite complete, 

 and closely resembles the female plumage, except for the wings, 

 which, of course, are molted only once in August, and for the breast 

 and belly, which remains largely brown. The molt into the eclipse 

 begins about the 1st of July and the change is very rapid. The molt 

 out of the eclipse in the fall is more protracted; it sometimes does 

 not begin until the middle of October and is not complete until 

 December or later. Adult males can always be recognized by the 

 wings. 



Food.— In feeding the shoveller uses its highly specialized bill to 

 advantage. All the surface-feeding ducks have the edges of the 

 upper and lower mandibles more or less well supplied with rows of 

 comblike teeth or lamellae through which the water and mud is 

 sifted to obtain food; in some species these are somewhat rudimen- 

 tary, but in the shoveller they reach their highest development because 

 the shoveller is more essentially a surface feeder than any other duck, 

 dabbling along the surface to sift out what small particles of food it 

 can find, shovelling in the soft muddy shallows and straining out its 

 food much after the manner of a right Avhale. The tongue, the roof 

 of the mouth, and the soft edges of the broad bill are all well supplied 

 with sensitive nerves of touch and taste, which helps the bird to 

 retain what it wants to eat and to reject worthless material. The 

 shoveller seldom tips up to feed by semi-immersion, but paddles 

 quickly along, skimming the surface, with its head half submerged 

 so that whatever is found is taken into the mouth, tasted by the 

 sensitive tongue, and sifted out through the pectinated bristles of 

 the bill if not wanted. 



Millais (1902) relates the following incident to illustrate the activity 

 of the shoveller in feeding: 



To the observer who sees the shoveler casually by day lie appears to be somewhat 

 of a lethargic nature; but, when he cares to do so, he can move faster on the water 

 than any of the fresh-water ducks. I have watched with pleasure the wonderful 

 sight, calculation, and quickness of a male shoveler that I once kept in confinement 

 on a small marshy pond at Fort George. About the last week in April a certain 

 water insect, whose name I do not know, would "rise ' ' from the mud below to the sur- 

 face of the pool only to be captured by the shoveler, who, rushing at full speed along 

 the water, snapped up the beetle the moment it came to the surface. How it could 

 see the insect in the act of rising I could never make out, lor it was inWsible to me 

 standing on the bank above, and I could only just catch a glimpse of it as the shov- 

 eler reached his prey and dexterously caught the beetle as it darted away again. 

 After each capture the duck retired to tlie side of the pool again and there awaited 

 the next rise-commonly about 25 feet away. While thus occupied he seemed to be 

 in a high state of tension; the feathers are closely drawn up and he kept liis neck 

 working backwards and forwards, in preparation, as it were, for the next spring, 



