166 BULLETIN 195, UNITED STATES NATIONAL MUSEUM 



few feet, slie works with feverish activity and with a directness of flight that 

 can scarcely escape the observation of even an untrained eye. 



Nevertheless, this period of construction is frequently interrupted by flights 

 to the beach or along the cliff in search of insects, or for a period of song on some 

 lofty point, or she too may dash out of sight far up the coast to return after 

 a period of from 5 to 30 minutes. 



Another favorable time for the location of the nest is during the incubation 

 period. Four nests under observation showed that the female remains upon 

 the eggs, whatever the character of the day or the stage of incubation, for a 

 period ranging from IS to 21 minutes. She then feeds from 2 to 5 minutes. 

 Here also her flight is relatively direct, in marked contrast to her usual journey 

 along the cliffs, and is unmistakable after a brief experience. The recorded 

 habits of several other birds indicate a fairly definite daily program during the 

 breeding season, but, so far as I know, none are so timed to the minute as the 

 Alaska Wren. 



All of the nests discovered in 1918 were in the faces of cliffs anywhere from 

 25 to 100 feet in height, and were placed at elevations varying from 8 to 100 

 feet. The spot chosen may be a crevice between shattered blocks of rock, or 

 in a small blowhole in the ancient lava flow, or, more frequently, underneath 

 banks of moss where rain and frost have excavated cavities of tidy size. In 

 three instances the nesting site had been chosen the year before, the new nest 

 being built upon the remains of the old one. In my experience the nest is never 

 hidden far beneath the general surface of the clitf. Of 12 nests described in 

 my field notes 4 were plainly visible, while the others were merely concealed 

 by an overhanging fringe of grass or moss or by a few small shattered scales of 

 rock. Four other nests were placed in cracks at a considerable elevation and 

 in overhanging cliffs that effectually prevented a close examination. 



The nest of the Alaska Wren is indeed a work of art, with the materials com- 

 posing it bearing a definite relation to the nature of its surroundings. Generally 

 speaking, it is a globular, more or less bulky affair with the entrance at one side. 

 When situated in a lava bubble or in cavities where the adjacent rock is rela- 

 tively dry, it usually consists of an external sheath of moss, thick or thin, ac- 

 cording to the size of the space to be filled. Where the soil inclines to be soggy 

 the roof alone is built of moss (at least in three instances) to absorb tlie moisture 

 and prevent its precipitation upon the sitting female. Farther down, at the 

 sides of the nest, it rests upon a meshwork of grass and roots that not only 

 drains away the water from above, but permits of rapid drying. To determine 

 the correctness of this theory a nest of this type was brought in from the field, 

 and was left overnight under the slow drip from a water tap. The next morning 

 the mossy roof was soaked and the grassy base adrip, but not a drop of water 

 had made its way into the interior. * * * 



The lining of the nest forms a heavy feltwork of which delicate roots and fine 

 filamentous lichen form the chief constituents. With these are usually associated 

 the feathers of the least auklet (and other birds to a less degree), fox hairs, and 

 in late years, the hair of the reindeer. 



There is a nest in the Thayer collection in Cambridge, which I 

 have examined. It was collected by E. C. Crompton on St. George 

 Island on May 20, 1922, taken from a crevice in the rocks of a cliff 

 on the seashore, about 20 feet up. It is quite bulky, being made 

 mainly of dry grasses and weed stems, mixed with green mosses and 

 lichens and a few feathers; it is lined with small feathers and very 

 fine white hairs. These white hairs were probably from the bleached- 



