'883-] Brewster oh an afparently New Gull. 2IQ 



feathers, where it occupies a longer space on the outer webs, 

 and on the second primary forms a complete sub-terminal bar. 



In Mr. Welch's example the fifth as well as the second primary 

 has a perfect sub-terminal bar, and the sixth shows an inter- 

 rupted one ; while the slate spreads over the greater part of the 

 webs of the first three feathers, except terminally. This exten- 

 sion of the dark color restricts the white spaces at the ends of 

 the second, third, fourth and fifth primaries to rounded apical 

 spots which resemble those of glaucescens. There is a further 

 approach to glaucescens in the unusually deep shade of the 

 mantle and the bluish cast of many of the light areas on the 

 primaries, but the mantle is still much lighter than in any speci- 

 men oi glaucescens which I have seen. 



In many respects L. kumlieni bears a curiously close resem- 

 blance to L. leucopterzis. It is of about the same size and 

 proportions, and the shape of the bill is similar, while several 

 of the specimens before me are positively identical in general 

 coloring. The only tangible point of difference seems to be 

 that of the peculiar wing-markings of ku7nlie7ii. This, of 

 course, is conclusive, but it is a matter of opinion whether it 

 indicates a stronger affinity with glaucescens. Welch's bird 

 certainly approaches glaticesceits., and a large series may estab- 

 lish a complete intergradation ; but, on the other hand, my 

 light extreme (the type) suggests a similar transition into 

 leucopterus. Were it not for obvious considerations I should 

 suspect that the bird might be a hybrid between glaucescens and 

 leucopterus. This, however, is highly improbable, and the 

 most rational conclusion seems to be that it is a distinct species, 

 intermediate between leucopterus and glatccescens^ but on the 

 whole perhaps more nearly allied to the latter, to which it bears 

 about the same relation that leucopterus does to glaucus. All 

 four species are evidently very closely related and form a group 

 of high northern distribution. The range of ktimlieni can be 

 only conjectured at present, but the evidence indicates that it is 

 probably confined to the eastern, and perhaps also northern, 

 shores of the continent, where it replaces true glazicescens., which 

 must be once more restricted to Pacific waters. Now that three 

 examples are known to have been taken near the eastern corner 

 of New England, it should be carefully sought all along our 

 seaboard, for it doubtless occurs there with some regularity in 

 winter, at least during severe seasons. 



