304 Phillips, The American Black Ducks. [july 



in his specimens is a form of variation which along with some others 

 he says he cannot account for. 



Aside, then, from certain other curious variations, we have in 

 this species male birds with Mallard affinities readily distinguished 

 from the females, a juvenile plumage like the females (Rothschild) 

 very likely a partial eclipse plumage, — in which the males lose 

 some of their characteristics, — and almost certainly some changes 

 due to age. The lack of accurate dates makes any assumptions 

 on this last point merely guesswork. A good deal of the difference 

 in color of the lower parts in this series is, I believe, due to staining 

 during life.^ 



I should like to point out what is at once the most noticeable 

 variation, and one that is certainly separated from any question 

 of age. This has to do with speculum bars. 



In the males the first band anterior to the speculum is always 

 black, but next to this there is a wide range of color difference. In 

 one case there is a broad white band, in two cases a narrow white 

 band, in two other cases a very indistinct buffy band, and in one 

 case no band other than the black one, which here is very finely 

 tipped with whitish. 



In the females {A. aberti included) the anterior margin of the 

 speculum is, in four specimens, bordered by a black band formed 

 by black tipped greater coverts, next comes a band of buff color 

 which shows a marked variation. In one case it is almost indis- 

 tinguishable, while in three cases it varies from a very light buff to 

 a russet color. In the fifth case. No. 113450, as Dr. Stejneger has 

 described, the normal white bar is replaced by a grey one, while a 

 new bar of pure white is interposed between the black bar and the 

 speculum, an arrangement entirely unique and non-mallard like. 

 The posterior end of the speculum appears to be always uniform, 

 it is bordered by two bars, an inner black and an outer white. 

 Thus we see a general tendency to variation anterior to the specu- 

 lum, as was previously pointed out for A. diazi. There are several 

 other minor points of interest about A. ivyviUiana, one of which is 

 a slight tendency to albinism in at least three specimens. 



1 cf. Wilson Bulletin, XXI, p. 221. 



