LIFE HISTORIES OF NORTH AMERICAN WILD FOWL. 137 



moist but not actually wet, and where the grass grew thick and lux- 

 uriantly. The nest was well hidden in the thick, green grass, so that 

 we never should have found it if we had not flushed the bird within 

 10 feet of us. It was merely a depression in the ground, well lined 

 with dry grasses, and sparingly lined with gray down around the eggs: 

 more down would probably have been added as incubation advanced. 

 The 10 eggs which it contained were perfectly fresh when collected 

 on June 3. 



The second nest was found on June 7 while driving across the 

 prairie in Nelson County. We had stopped to explore an extensive 

 tract of low "badger brush," looking for the nest of a pair of short- 

 eared owls which were flying about, as if interested in the locality. 

 We were apparently a long distance from any water, and while re- 

 turning to our wagon over a high dry knoll, flushed the duck from 

 her nest, v.hich w^as only partially concealed in the short prairie grass. 

 The slight hollow in the ground was lined with dead grasses and a 

 plentiful supply of down. It contained 11 eggs which were too far 

 advanced in incubation to save. Although the shoveller frequently 

 breeds in open and exposed situations at a long distance from water, 

 I think it prefers to nest in the rank grass around the boggy edges 

 of a slough or pond. 



In southwestern Saskatchewan in 1905 and 1906 we found shovellers 

 everyvv-here abundant, breeding on the islands, on the meadows near 

 the lakes, and on the prairies. On that wonderful duck island in 

 Crane Lake, on June 17, 1905, we found 7 nests of the shoveller — 2 

 v/ith 8 eggs, 1 with 9, 2 with 10, and 2 with 11 ; the nests were located 

 in the long grass and under rosebushes, scattered indiscriminately 

 among the nests of mallards, gadwalls, baldpates, green-winged and 

 blue-winged teals, pintails, and lesser scaup ducks; this island has 

 been more fully described under the gadwall. The nests v/ere very 

 much hke those of the other ducks, hollows scooped out in the 

 ground, sparingly lined with dry grass and weeds and surrounded by 

 a rim of down; as incubation advances the supph" of down increases 

 until there is enough to cover the eggs when the duck leaves the nest. 



I believe that the above-described nests illustrate the normal nest- 

 ing habits of the shoveller, but Mr. Edward Arnold (1894) records a 

 nest which "was built in a heavy patch of scrub poplars," in Mani- 

 toba. Mr. W. Otto Emerson (1901) thus describes a nest which he 

 found in California in an exceedingly exposed situation in a salt marsh : 



After working over the marsh for several hours 1 started back aiui when half way 

 across I again saw a pair of ducks headed inland, but thought nothing of it until a 

 single duck started up 10 feet from me and 300 yards from the mainland. On going 

 to the spot there lay a nost in open sight on the bare ground among the saltweed. 

 It was not over 4 inches off the ground and contained 14 eggs. The nest was com- 

 posed of dry stems of the saltweed, lined with down and a few feathers from the parent 

 bird, and measured 14 inches across the top with a depth of 5 inches. 



