AFRICAN YELLOW-BILLED DUCK 
115 
feet, and bill are black (Blaauw, 1919). Rogeron (1903) notes that it would be impossible to distin- 
guish them from young Mallards, except for the bill, which very soon takes on a yellow tint. The 
many specimens which I saw in the Leyden Museum and also in the British Museum led me to think 
it almost impossible to tell them from Mallards. In some, the dark ocular streak does not go forward 
of the eye and in others the yellow of the under side is very sharply defined from the brown of the 
upper side. But there are so many little variations in the down plumage of ducks that comparison of 
a few specimens is often misleading. 
Remarks on Proposed Races: The material which I have seen is insufficient to decide the validity 
of the separation which has been proposed. A northern form was recognized by Blyth (1856), but 
was not considered valid by Salvadori (1895). It was re-instated by Neumann (1904). In view of the 
fact that speculum-color appears to be a character of very doubtful value, and because birds with 
green specula have been noted in South Africa (Sassi, 191£), I have thought it best not to recognize 
A. u. rupfelli from northeastern Africa (Shoa). 
DISTRIBUTION 
Abyssinia 
Like many tropical species the Yellow-billed Duck is only locally migratory, and is found throughout 
all parts of its range at very different seasons. It is primarily a South African species, and south of 
15° south latitude or at least 20° south latitude it is the commonest duck. In the east, 
however, the species is found as far north as Abyssinia where it is common and breeds, 
especially in the highlands of the south. Von Heuglin (1873) found it as far north as Lake Tana; 
Salvadori (1884, 1888) has repeatedly recorded it from Lakes Addo and Cialalaka and other local- 
ities in Shoa, as well as from Auasc on the Italo-Ethiopian frontier (Salvadori, 1912). The U.S. 
National Museum has specimens from the Amssi Plateau. Blanford (1870), and Ogilvie-Grant and 
Reid (1901) have recorded it from the highlands: A. D. Smith (Sharpe, 1895) and Peel (1900) found 
it at Sheik Mohammed in western Somaliland, and von Erlanger (1905) states that it is especially 
common and breeds between Ginir and Addis-Ababa and on the Maki River as far as Addis-Ababa. 
But the species seems in this region to be confined to Abyssinia. I find no records for the Sudan nor 
for either British or Italian Somaliland proper. 
Farther south it is common and presumably breeds on the Toro Lakes, Uganda (V. van Someren, 
1916) ; and is common also in British East Africa (Horsbrugh, 1912; etc.) whence it has been recorded 
for Lake Naivasha at altitudes up to 2000 meters and where it was found at Kagio, Kutu and at 
Kenia, breeding at altitudes of 2700 meters (Lonnberg, 1912). It is apparently com- Africa 
mon also in former German East Africa where it has been found at Iringa (Reiche- 
now, 1898), while from Victoria Nyanza it has been recorded by Reichenow (1894). Sassi (1912) 
found it in the Urundi region and Emin Pasha (1891) met with it at Bukoba on the west shore 
of the lake. Johnston (1886) has recorded the species for KOimandjaro where it was found as 
high up as 11,000 feet. What the status of the species may be to the west of the lake region 
it is very difficult to say. Very likely it occurs throughout Congo Free State, for Johnston 
(1884) states that he found it at Stanley Pool and at suitable places far up the Congo ^QjjgQ 
River, while Neave (1910) has recorded it for the Katanga, Bangweolo and Kalung- 
wisi regions, up to 4200 feet, in the southeastern Congo. In Portuguese West Africa it is, how- 
ever, a rare bird, having to my knowledge been recorded only from Huilla and Caconda in the 
interior (Barboza de Bocage, 1877-81). Andersson (1872) did not meet with it in former German 
West Africa and I believe it has never yet been found there; so that at best the species must be 
regarded as very rare on the west coast north to the Congo. Rochebrune’s (1883-85) statement 
that it is rare in Senegambia and that specimens have been taken north to the 
lower Gambia, requires further confirmation. In Rhodesia and the other British South 
African colonies it is, however, very abundant (Layard, 1875-84). Holub and von Pelzeln (1882) 
