34 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 64. 



middle toe, 4&-53.5 (51.6). The four females measures: Wing, 231- 

 256 (249.7) ; tail, 92-100 (97.2) ; culmen, 48.5-52 (50.2) ; tarsus, 

 40.5-45.5 (42.2) ; middle toe, 46-55.5 (51.6). Three specimens from 

 Australia (only one sexed and that a female) measure: Wing, 275- 

 255 (263.3) ; tail, 99.5-108.5 (103) ; culmen, 53-58.5 (54.8). 



Anas supercilosa felewensis is still smaller than the Celebes form, 

 the buff of the throat is lighter, the auriculars and sides of neck 

 more heavily streaked, and there are other differences. 



The downj^ young taken at Koelawi, February 2, may be described 

 as foUows: Above sepia, darker on the rum,p; sides of face and 

 lower parts, cream-buff, deepening on sides of face ; an obscure band 

 across chest and a very narrow line down the center of breast, cin- 

 namon-buff; a dark line from bill under and through the eye to 

 the najDc; a rictal spot and another on the auriculars, of the color 

 of the back; superciliary stripe, creamy buff; a line across hinder 

 border of wing, a small streak on the back on each side opposite the 

 wing, and another small streak on each side of rump, cream-buff; 

 flanks and crissum, a little lighter than the back. 



For a fuller discussion of the forms of this duck the reader is 

 referred to Mathew's Birds of Australia,^^ where references to the 

 pertinent literature will be found. 



64. NETTION GIBBERIFRONS GIBBERIFRONS (S. Muller). 



A good series of both sexes from: Toli Toll, November 28-De- 

 cember 12, 1914; Kampa, February 14, 15, 1915; Tandjong Penjoe, 

 February 16, 1915; Winatoe, January 21, 1917; Koelawi, February 

 23, 1917; Rano Lindoe, February 24-March 13, 1917; Gimpoe, 

 August 4, 1917. 



The above birds when laid out in series show quite a little varia- 

 tion. The specimens taken in January, February, and the early 

 part of March are much lighter below, without any chestnut wash on 

 the breast, and the throats are much lighter, also, almost white. The 

 birds taken in the latter part of March and December have the 

 underparts strongly washed with chestnut and the throats tinged 

 with rusty, but in my opinion this difference in color is due to 

 stain caused by iron in the water. The males have the foreheads 

 considerably swollen, much more pronounced in some specimens than 

 in others; it is very noticeable in the skeleton, but poorly developed 

 or almost entirely absent in the females. 



A pair in the United States National Museum from Java are con- 

 siderably darker on the back, wings, scapulai^, and tail than any of 

 our Celebes specimens and probably represent a distinct form. 



85 Vol. 4, pt. 2, 1915, pp. 85-94. 



