BIRDS OF ETHIOPIA AND KENYA COLONY 85 



insignis, purpurascens, and lacuum. However, the range of lacuum 

 is not correctly stated by Sclater or by Neumann.®*^ According to 

 both of these writers, it is restricted to the southern Shoan lake 

 region and the Omo district. The present specimens from Hawash 

 River and Serre, however, are clearly referable to lacuum and not to 

 the north Ethiopian race leucomelas^ of which, unfortunately, I have 

 not seen any material. Tlie latter race is said to differ from lacuum 

 only in being smaller. Neumann gives the wing length of leucomelas 

 as 75 to 84 mm, that of lacuum as 85 to 95 mm. The female from 

 Hawash River has a wing measuring 85 mm ; the females from Serre, 

 86 and 88 mm, respectively. Furthermore, van Someren ®^ states that 

 his series of lacuum, from Turkwell south to Kakamegos presents 

 wing lengths of from 83 to 91 mm, slightly less than Neumann's 

 figures and in better agreement with the present series. It appears 

 that leucomelas occurs in Eritrea, Sennar, and the highlands of north- 

 ern Ethiopia south to northern Shoa, but not to eastern Ethiopia, 

 while lacuum occurs in the lower elevations of the Hawash valley, 

 southwest (through Ennia-Gallaland) to southern Shoa, the Omo 

 region. Lake Stef anie, and northwestern Uganda. I wonder whether 

 van Someren's Kakamega birds are really lacuum; on geographic 

 grounds they would seem to be purpurascens, the form occurring on 

 Mount Elgon and across Uganda to the eastern Congo, southern 

 Sudan, and to northern Cameroon. 



In northeastern Africa three races of this bird occur, as follows : 



1. P. n. leucomslas : According to Sclater, this form is found only 

 in Eritrea, Sennar, and the northern highlands of Ethiopia. Lynes,^^ 

 however, refers his Darfur birds to this race, and Bannerman and 

 Bates ^^ list two specimens of leucomelas from between Ibi and 

 Takum, and from east of Bauchi, in Nigeria. I have seen no mate- 

 rial from Darfur or Nigeria and assume that Sclater must have 

 examined these birds and found them to belong to the race pur- 

 purascens. 



2. P. n. lacuum,: This form has been discussed above. Its range 

 is from the middle stretches of the Hawash River to southern Shoa, 

 the Omo region, and northwestern Uganda. It resembles leucomelas 

 but is smaller. 



3. P. n. purpurascens : Mount Elgon, across Uganda to Ruwen- 

 zori, the eastern Uelle district of the Belgian Congo, north to INIon- 

 galla and the Upper Nile provinces of the Anglo-Egyptian Sudan, 

 thence west through the southern Sudan (north to Darfur) to 



»" Jouin. fiir Orn., 190G, p. 260. 

 »iNov. Zool., vol. 29, 1922, p. 205. 

 »2 Ibis, 1924, p. 719. 

 »3 Ibis, 1924, p. 250. 



106220—37 7 



