188 BULLETIN 113, UNITED STATES NATIONAL MUSEUM. 



Franz Josef Land, in August. The white lesser wing-coverts show 

 very plainly in this plate, but apparently these are molted before 

 October, as they do not show at all in Nelson's (1887) plate. 



Nelson's bird, taken October 10 at St. Michael, illustrates the 

 change into the first winter plumage, which differs from the fore- 

 going in having the under parts pure white and having lost all traces 

 of the white lesser wing-coverts, the ochraceous barring on the back 

 and the ochraceous tip of the tail. Some of this may be accounted for 

 by wear or fading, but the appearance of a few " mantle blue " feath- 

 ers in the back indicates a partial molt, including, at least, the back 

 and lesser wing-coverts. 



The young bird figured in P. H. Ray's (1885) report, a bird in its 

 first autumn, is considerably more advanced, as it has a pink breast 

 and a blue mantle, the wings being practically the same as in Nel- 

 son's (1887) bird. This specimen illustrates the first winter plumage, 

 which is worn with little change until spring. A partial molt, in- 

 volving nearly, if not quite, all the contour feathers and the tail, 

 takes place in the spring. This first nuptial plumage is illustrated 

 by a bird in the author's collection, taken on the Kolyma River on 

 June 2. In this the wings, including all the coverts, are as in the 

 first winter plumage. The remainder of the plumage is fresh; the 

 mantle is clear blue; the underparts are decidedly rosy; the tail is 

 pure white ; and the black neck ring is indicated by scattering black 

 feathers. At a complete postnuptial molt, the following summer, 

 the young bird would probably assume the adult winter plumage. 



Adults apparently have a complete postnuptial molt and an in- 

 complete prenuptial molt of the contour feathers. The black ring 

 on the neck is characteristic of the nuptial plumage, but the pink 

 underparts and the pure white tail are present at all seasons. 



Fob A. — Buturlin (1906) noted the Ross's gulls chasing insects, and 

 the stomachs which he examined contained " only fragments of 

 coleoptera, gnats, and other insects;" therefore, on their breeding 

 grounds, at least, their food seems to be wholly insectivorous. 



Behavior. — Although he first noted the flight of this species as 

 " more Fulmar-like," Buturlin (1906) finally concluded that it 



was really much more tern : like * * *. The rosy gull swims easily, and 

 sometimes I saw it taking a regular bath. It dipped its head tinder, while 

 sitting deep in the ice-cold lake, and, throwing the water over its back, moved 

 its wings quickly below the surface, holding itself almost clear, and threw 

 Itself forcibly head downward into the water. Once a rosy gull flew over the 

 surface of the lake with a cry of " carvac-wa " and took up water with its 

 beak on the wing, as swallows do, but subsequently it settled on the surface for 

 some two or three seconds without folding its wings, which were elevated over 

 the back, and drank after the usual fashion. 



The note of RhodostetMa is peculiar, being high and more melodious than 

 that of gulls in general, and very variable. The cries that I most often heard 



