1883.] PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 99 



^ A few words in regard to the true relationship of tin- species may be 

 in order first, as the original description in a general way only refers it 

 to " the section of true Anas, which embraces A. obscura of North -Ymeri 

 ca, A. superciliosa of Australia, and other species." From the synonymy 

 as given above it might be supposed that the relationship of the present 

 species would be with A. superciliosa, the habitat of which is given as 

 embracing nearly entire Polynesia, besides Australia and New Zea. 

 laud, but in point of fact it has nothing to do with the Polynesian spe 

 cies, as its closest allies are found on the American continent just oppo- 

 site the Hawaiian Islands; or, to be more specific, A. wyvilliana is very 

 closely related to A. aberti Pidgw. of northwestern Mexico, as already 

 pointed out by Mr. Ridgway (Proc. U. S. Nat. Mus., i, 1878, p 251) 

 ; So close is the similarity that I am unable to distinguish No' 113450 

 A. wyvilliana ( 9 ), from the type of A. aberti ( 9 ) except by the larger size 

 and the total absence even of an indication of supraocular or trausoc- 

 ular stripes; the former specimen is also distinguished by white tips to 

 the greater wing coverts, forming a white anterior border to the o TOe n 

 speculum, but this is wanting in the other specimens as well as in the 

 type of A. aberti. 



On the whole, the five specimens of Anas wyvilliana before me 

 though agreeing rather closely in general coloration, present a number 

 of extraordinary variations truly perplexing. 



Nos. 113447 and 113449 I take to be correctly determined as male* 

 for they have the central tail-feathers turned up at the end in a manner 

 similar to that of the male Anas boschas* Thev also differ from the 

 other three specimens before me in having the smaller upper wing- 

 coverts nearly uniform « hair brown" (Pidgw., Norn. Col., pi. iii „ 19) 

 'with but faint grayish margins, while the other three have' these 

 feathers distinctly and, more or less, broadly margined with cinnamon 

 The two males, furthermore, have a slightly stronger greenish gloss on 

 the occiput and upper hind neck, but here the agreement between them 

 as compared with the three other specimens, comes to an end. 



No. 113449 has a small black bill, with an ill-defined yellowish space 

 on each side of the upper nail, and a somewhat asymmetrical longi- 

 .tudinal mark of similar color on the middle of the column, and a yellow- 

 ish area separating the under nail from the dark basal half of the lower 

 mandible, while in No. 113417 the bill is pale brownish, becoming light 

 plumbeous towards the base, with a triangular black mark on the lower 

 basal angle of the upper mandible exactly as in Anas fulvigula Pidgw. 

 [n No. ll3449jhe_to p of the he ad^occiput, and upper neck are nearly 



* The occurrence oM^n^^^ is exceedingly 



.nteresting, as they are said to be entirely wanting in Anas obscura ami .1,,. still more 

 Closely allied A. fulvigula from Florida. That these curious feathers are not indicated 

 u the original description nor in the figure (P. Z. S., 1878, and ' Challenger' Rep.) is 

 'asily explained by the tact that the type was collected ,u August, at a time when 

 nese feathers are normal even in A. boschas. Our specimens are apparently winter 



