duo Brooks, Notes on American Ducks. LJuly 



3. Wider gray pectoral band. Probably a sound distinction 

 but one which is dependent, in the skin, on a uniform method of 

 make-up. 



4. Shape and proportions of bill. The more tapering bill of 

 islandica is a thoroughly good distinction, but as given by all 

 authorities it is a comparative one, just as is " bill more goose-like." 



How can a man who has only one Golden-eye tell whether the 

 bill is tapering when no measurements are given, or that it is 

 goose-like if he has no example of the other species to compare 

 with it? Ridgway's 'Manual' however has a definite formula 

 as the best distinction between the two species. 



" A 1 . Height of upper mandible at base, measured from point of frontal 

 angle to nearest point on cutting edge, less than distance from anterior 

 edge of loral feathering to anterior end of nostril, and usually little if any- 

 greater than distance from latter point to tip of upper mandible 



G. clangula. 



151. G. clangula americana." 

 "A 2 . Height of upper mandible at base, measured from extremity of 



frontal angle to nearest point on cutting edge, equal to distance from an- 

 terior point of loral feathering to anterior end of nostril, and much greater 

 than from latter point to tip of upper mandible 



152. G. islandica." 



I have carefully tested this with the following results: 



If the first measurement is taken by placing orte point of the 

 dividers on the frontal angle, and the other on the cutting edge, 

 i. c, the chord of the distance around the bill, the results are 

 wrong in one-fourth of my specimens of islandica, including 

 one adult male, and wrong in two-thirds of my specimens of 

 americana including two adult males. 



If this measurement is taken by holding one point of the dividers 

 level with the frontal angle in line immediately above the nearest 

 point of the cutting edge, i. e., the actual height of the former 

 above the latter, the results are hopelessly out in nearly all my 

 islandica, but correct for all but one of my americana. 



If the first portion of the proposition is taken, eliminating 

 the distance from nostril to end of bill, measurements taken by 

 first method, the results are correct in all my americana, but 

 wrong in four out of eleven islandica. So this distinction, which 

 I had great hopes of, proved a "no thoroughfare." 



