100 BIRDS OF HAWAIIAN ISLANDS. 



uniform dull blackish brown, with a greenish -loss on occiput and neck, 

 while the fore-neck has a whitish ground color striped with blackish; 

 iu No. L13447, on the other hand, these parts arc colored as in the fe- 

 male, pale cinnamon striped with dusky underneath, and dusky striped 

 with pale cinnamon above, though with a stronger greenish -loss than 

 in the females. No. 113440 has the anterior white wing band nearly 

 twice as broad as the corresponding one in 113447. Lower wiugcov- 

 erts and lower tad-coverts show great difference in the two specimens, 

 No. 113447 agreeing in the coloration of these parts with the other three 

 specimens in having the entire lining of the wing, including the axillar- 

 ies, pure white, and in having the lower tad coverts whitish, more or less 

 tinged with tawny and ciunamou rufous, and spotted or barred with 

 dusky; No. 113449, however, has all the small under wing-coverts 

 brownish may. with broad, pale margins, and the under tail-coverts are 

 deep cinnamon rufous, heavily blotched with black, some of the middle 

 ones being entirely black, with a faint greenish gloss. Also among the 

 upper tail coverts of the same specimen there are a few uniform 

 greenish-black feathers. 



No. 113418 is said to be a male by the collector, but, if so, it is proba- 

 bly a younger bird than the two already spoken of, for not only are 

 the central tail-feathers perfectly straight, but the smaller upper wing- 

 coverts are broadly margined with ciunamou, while several of them iu 

 addition have a central pale heart-shaped mark distally circumscribed 

 with black; the nape has a greenish gloss, but slightly fainter than iu 

 the foregoing specimens, but the lower neck and upper breast are mure 

 rufous than any of the specimens at hand; the upper white wing band 

 is narrow and considerably clouded with gray. Bill as in No. 11341.. 



The last two specimens are said to be females, and are probably cor-J 

 rectly sexed, but they present great differences inter se. No. 21310 ( t he 

 same one described by Mr. Kidgway, loc. eit., though the number by a 

 misprint is given as 20319) seems to be the more normal of the two,, 

 but the upper white wing-bar is nearly as obsolete as in the type of A. 

 abcrti, it being only represented by a narrow, ill-defined, pale drab- 

 gray band, and round the eye there is a somewhat irregular ring of 

 white feathers. The latter feature, however, is probably purely albij 

 nistic, there being also a few small white feathers here and there on the 

 fare, and none of the other specimens show even an indication of au 

 eye ring. The bill appears to be like that of Nos. 113447 and 11344$ 

 The tawny margins to the smaller upper wing-coverts are narrower than 

 in the foregoing and following specimens.* No. 113450, the last of 

 Knudsen's birds, is particularly interesting and aberrant. While al' 

 the other specimens have the throat streaked with dusky, this one hj 

 the chin, throat, upper fore-neck and the loral apex uuitorm, pale vn 



•The specimen just described was probably obtained in March, W5fi, at Hilo, Isld 

 of Hawaii, consequently in the identical locality vvhence^eame the fcype of A. in/n 

 liana. 



