ANCIENT MURRELET 135 



thing goes. Fortunately it is impossible to find all the nests, or kill all the 

 birds, so enough remain to stock the island again another season. 



By no means every island in this vicinity is occupied by murrelets. Within 

 400 yards of the one of which I write is another of about the same size and 

 topography, but strange to say, no murrelets are found on it, although there 

 are two or three small colonies of auklets, the remainder of the island being 

 given over to Leach's petrels. Again, on two other small islands, also near to- 

 gether, each containing about a couple of acres, and in every way alike, one is 

 given over entirely to auklets, while on the other the murrelets have almost com- 

 plete control. These facts cause me to believe thai* the birds always return 

 to the island on which they have been reared. 



On June 23 our party returned to the island on which we first landed, and 

 found to our great satisfaction that the murrelets' eggs 4 were more plentiful 

 than on our former visit, and a few of them were taken. We also soon dis- 

 covered that they were not especially particular in the selection of a nesting site. 

 An abandoned burrow of Cassin's auklet, a dark crevice in cliffs, under large 

 broken rocks which had fallen from the latter, or under large tussocks of rank 

 grass, with which the higher portion of the island was covered, would answer 

 equally well. Under these almost solid bunches (the grass remaining from 

 several previous years), the murrelets would force their way, leaving only a 

 slight hole in the mass, which usually was very hard to detect. After once gain- 

 ing an entrance into this matted vegetation and working their way in for 2 or 

 3 feet, a shallow cavity about 5 inches in diameter and 2 or 3 inches deep, 

 was scratched out and this was nicely lined with blades of dry grass of last 

 year's growth, carried in from the outside, making a very neat and snug home, 

 in which the two beautiful eggs comprising a set, were deposited. Some of their 

 nests were found fully 200 yards from the water. In the other situations men- 

 tioned little and often no nest is made, and the eggs are deposited on the bare 

 rocks, in the soft sand, or on the wet, muddy soil. I even took several sets on 

 the bare ice at the bottom of some auklet's burrow, the ground being still 

 frozen, immediately beneath the grass and moss on July 3, when I left the 

 island. 



The setting bird will sometimes leave the nest when danger threatens, but 

 it will frequently allow itself to be taken from the eggs, and when brought 

 to light it will screech, scratch, and bite with vigor. When released they can 

 not fly unless thrown into the air, and will then often fall back to earth. One 

 evening, just at dusk, I was crouched in the grass waiting for a shot at a 

 Peale's falcon (Falco peregrinus pealei), who made regular trips to the island 

 to prey on the auklets and murrelets, when I heard a very low but rather 

 shrill whistle. Turning my attention to the spot from which it seemed to come, 

 I listened; presently I heard it again, but was still unable to locate the bird, 

 which I afterwards found to be a murrelet. Subsequent observations proved 

 that this was a call note uttered just about the time the setting bird expected 

 the return of its mate, and was evidently uttered to attract his or her attention, 

 for as far as my observations went, they, like the auklets, exchange places 

 nightly, and while one attends to the home cares, the other is usually a number 

 of miles out at sea on the feeding grounds. This call note is the only one I could 

 attribute to this species while on land, and so ventriloquial are their powers, 

 that in only two instances did I succeed in locating the nest from the sound. 

 While out at sea, the ancient murrelet utters a peculiar piping whistle entirely 

 different from the one uttered while on the nest. 



Two eggs are laid to a set, the second is deposited after an interval of two 

 or three days, and frequently three or four days elapse before incubation be- 



