no spring nioult^ but the green colour gives way to a beautiful steel- or violet-blue. This is true 

 of the males and of the old females; but we fancy that the latter take much longer in arriving 

 at the full metallic plumage than do the males, which probably arrive at perfection in their first 

 spring. ]\Ir. Brewster has drawn attention to the white fringes of the secondaries^ which he 

 believes to be an accompaniment of the winter plumage of the old birds as well as of the young,. 

 and this seems to be confirmed by the examination of the specimens in the Salvin-Godmau 

 collection ; but these white tips are speedily east when the spring plumage is complete and the 

 birds are in their summer haunts. After the autumn moult the upper tail-coverts are tipped 

 with ashy or whitish margins, and, curiously enough, these white edges sometimes accompany 

 the bird on its return journey in summer, and are even occasionally developed in a high 

 degree. 



Young. Very different from the adults and without metallic gloss. General colour above sooty or 

 mouse-brown ; the wing-eoverts and quills like the back ; lores blackish ; ear-coverts like the 

 head; cheeks, throat, and under surface of body white; fore neck and chest pale mouse-brown, 

 darker on the sides of the chest and flanks: " bill black; rictus and interior of mouth yellow ; 

 feet dark livid sepia or light pink; iris brown" [R. Rldgwaij). 



A good series of j'oung birds is contained in the Henshaw collection, the earliest bearing the date of 

 July 4th. Mr. Brewster observes : — " The first plumage is worn much longer than in most 

 birds, and the autumnal dress very slowly acquired, the metallic tinted feathers appearing one or 

 two at a time. The remiges are also moulted by the young as well as by the adult, and both in 

 the autumnal plumage have the last pair of secondaries broadly tipped with pure white. This 

 remarkable feature, so far as the specimens at hand go to show, is entirely characteristic of this 

 plumage." Again, in the autumn of 1879, he continues:- — "From investigation of material 

 collected during the past season I find that the change takes place from about the middle to the 

 last of September. Six specimens shot at Concord, Mass., October 16th, 1878, have all acquired 

 the full autumnal dress." 



In the Henshaw collection, however, are specimens which are commencing to moult on the 

 14th and the 27th of August, and an old female killed in the same month and at the same jilacc 

 (Sing-Sing, N.Y.) has almost completely donned its winter plumage, including the white-tipped 

 secondaries; the longest primaries have not yet been shed. One male, killed at Cambridge, 

 ilass., on the 30th of September, and marked by ]Mr. Henshaw as a male in its first winter 

 plumage, is entirely green above, with well-marked white margins to the inner secondaries ;incl 

 upper tail-coverts, and has a shade of brown over the fore neck and chest. This is also seen in 

 young birds in the first plumage, but every sign of it disappears in the sjiring dress. It is pro- 

 bable that the earlier moulting birds are those of the earlier hatched broods. 



A female killed by Mr. Nelson at St. IMichaers, Alaska, on the .2 1th of ^lay is bri)wn above, with very 

 little gloss left; and this may prove that, as we have above suggested, the IVnudes do not ac(|uire 

 their full brilliancy of jjlumagc in the first spring. Certain it is tliat some f'emak's are iidly :is 

 brilliant as any males, but tiiey may take really two or three years to aiquire this perlett dress. 

 Or it may be that the wear and tear of the journey to the far north and the more rigorous 

 climate may cause an early abrasion of the metallic pluniage. 



Hab. Nearly the whole of North Anieriea, Init its exact northern r;uige is not yet determined. ^\■interiuL' 



in the Gulf States and in Central America, as far as (1 nateni:d;i, in liennnda, I lie l)ali:ini;is, and 



Cuba. 



■2 K -2 



