166 BULLETIN 113, UNITED STATES NATIONAL MUSEUM. 



eggs were fresh and some heavily incubated, showing that they were 

 probably laid by two birds. After collecting a few sets of eggs and 

 exposing a lot of plates we reluctantly came away, not having fired a 

 gun among the beautiful and confiding birds. We even refrained 

 from killing one which had become tangled in the reeds and was 

 easily caught. We visited this locality again the following year, 

 but were disappointed to find it entirely abandoned by the gulls, 

 which was probably due to the fact that the lake had been very dry 

 earlier in the season when they were beginning to nest. 



The colony described above was undoubtedly unusual, as this 

 species generally nests in a more open marsh in deeper water and 

 often in quite exposed situations in marshy lakes. In the three other 

 colonies that I have seen the nests have been floating in water which 

 was waist deep, or deeper, and many of the nests could not be reached 

 without a boat. Dr. Thomas S. Roberts (1900) describes a typical 

 deep-water colony in his account of the nesting habits of this species 

 at Heron Lake, Minnesota, in 1899. He says : 



At a distance of about an eighth of a mile from the marshy, reed-grown shore, 

 the little floating mounds dotted thickly a great crescent-shaped area some 

 three-fourths of a mile in length by 300 or 400 yards in the widest part. The 

 nests were irregularly distributed. In some places there were many close 

 together, and again they were scattered yards apart, while now and then there 

 were large spaces where there were none at all. 



Under ordinary conditions the water over all this area would have been 2 or 

 3, nowhere over 4 feet deep, with a thick growth of bullrushes (Scirpns) 

 standing well above the surface. But heavy rains had raised the lake until the 

 water was in many places fully 6 feet deep and only the tops of the tallest 

 rushes came into view ; thus changing a large part of the nesting ground from 

 a dense tangled bed of rushes into almost open water. Upon this condition 

 of things the birds, of course, had not reckoned when they chose the site, and 

 in consequence many of the nests were now torn from their moorings, having 

 been lifted by the rising water, and were unprotected save by the weak tops of 

 the submerged rushes. Thus free to drift, they were floating hither and 

 thither at the mercy of the winds, but, strange to say, this state of things did 

 not appear to greatly disconcert the owners. Here and there a number of nests 

 had caught against some firm anchorage, and, receiving new additions with 

 each favorable breeze, a windrow, or island, of these stray nests was soon 

 formed. Nest touching nest in this manner resulted in a promiscuous crowding 

 of families that must have tested the good nature and forbearance of the occu- 

 pants not a little, and probably led to some vagaries in the care of the young 

 described further on. A few nests had gone adrift entirely, and floating far out 

 into the open water had been abandoned. But luckily a considerable part of 

 the colony, wiser than their fellows, escaped this dire confusion of disaster as 

 the result of having located their nests where shallower water and stronger 

 growth of rushes provided protection and safe anchorage even when the flood 

 was at its height. From nest-building operations still in progress at the late 

 date of our visit (June 16) we inferred that a few at least of the gulls that had 

 lost their homes were reestablishing themselves in safer retreats farther back, 

 having perhaps learned a lesson against future similar mishaps. 



