The Terns of Weepecket Islands 17 



raided after June 28, and as incubation was quite well along 

 at that date, it seems hardly likely that a previous brood had 

 been raised, and there was no sign of young birds. A party 

 visiting the island from the Laboratory on July 10 reported 

 but a few nesting birds, and innumerable young, still in the 

 down, running about the island. On July 38 no nesting birds 

 were seen, and the young were rapidly maturing. August 7 

 found the island entirely deserted. This should not be taken 

 to mean that in a healthy, flourishing colony, free from inter- 

 ference, two broods are not raised during a season. 



The activity in the colony was incessant, and there was 

 hardly a time during the two days of my visit that one sec- 

 tion or another of the island was not in commotion. This 

 activity lasted well into the night, and those few birds which 

 had already hatched their young were bringing in fish as late 

 as 8 :20 in the evening, and as early as 3 -.28 in the morning, 

 so far as I could see. Most of these fish, by the way, are 

 caught around the neighboring islands, and are even brought 

 from the shore of the mainland. In the harbor at Woods 

 Hole the birds were watched diving for fish {Fundulus 

 heteroclitus and Ammodytes americanus largely), the birds 

 remaining completely under the water for a second or a trifle 

 longer. 



The fact that the island was so crowded with nests sug- 

 gested one or two little experiments, which were tried before 

 leaving, to see if the bird recognized its own nest, and the re- 

 sults seem to point to the conclusion that the bird returns to 

 the spot, rather than to the nest itself. The eggs from two 

 nests were interchanged, and the bird on returning, settled 

 without noticeable hesitation on the new set of eggs, though 

 her own were in another nest, fully exposed to her view, less 

 than three feet from her. In the case of a grass-lined nest 

 with a single egg, the rude lining was removed, and the com- 

 plement increased to four eggs by temporarily robbing neigh- 

 boring nests, and the bird did not hesitate over the remarka- 

 ble increase and change. The reverse of this proved equally 

 true ; a complement of three reduced to one did not seem to 



