318 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



and southern birds from a comparatively small area also vary, it is not 

 unreasonable to suggest that P. c. omoensis is not a good race. 



Dr. Hartert, in fact, is inclined to this view, but I am not in agreement with 

 this. I suggest that omoensis is a good race, and that possibly there is another 

 race inhabiting the south end of Rudolf and Bariugo districts, with characters 

 as given above. 



Recently, van Someren ^^ obtained seven more skins from Kaptirr, 

 Turkwell, which he compared with his previous series, and found 

 that the color of the hind crown and nape is variable. 



I have examined 15 adult birds from southern Ethiopia (Ourso, 

 Abaya Lakes, Bodessa, and Turturo) and Uganda, and find that the 

 color of the occiput and nape varies from neutral gray to a slate- 

 gray, with some plumbeous feathers mixed in. These all ought to be 

 omoensis, with dark, that is, slate-gray, napes. Other recent in- 

 vestigators have also found ^^omoensis''^ to vary. Thus, Stoneham " 

 notes that birds from Kitgum, Uganda, have much darker gray 

 napes than a specimen from Karamoja (a locality nearer to the Omo 

 drainage basin). He suggests that wear may account for the dark- 

 ness of the Kitgum birds. In this he appears to have hit upon the 

 correct solution. I have gone over my material and have found that 

 the birds in fresh plumage have paler napes, those in abraded condi- 

 tion, darker, more slate grayish. Although I have examined no birds 

 from Eritrea or northern Ethiopia, I feel confident that if due allow- 

 ance be made for the condition of the plumage, the alleged differ- 

 ences between cnstata and omoensis wdll disappear. 



With regard to the buffy-naped birds from south of Lake Rudolf, 

 a similar type of variation occurs in P. concinnatus, especially in 

 western Kordof an. Lynes ^^ described the Kordof an birds as dis- 

 tinct {ochracea), but later ^"^ he found the variation to be inconstant. 

 This seems to be the case with P. cristata as well. 



The species, then, contains three races, as follows : 



1. P. c. cristata: Eritrea, Ethiopia, except the eastern part of the 

 Harrar district and southern Gallaland, south through Shoa and 

 Arussiland and the Omo region to Uganda through Turkanaland 

 and northern Uganda to Kisingo and Kigomma, and to the north- 

 western part of Kenya Colony (Lake Rudolf to Lake Baringo and 

 to Mount Elgon). The last part of its range may be inhabited by 

 a recognizable race, but in spite of van Someren's assurance of the 

 maturity of his birds, the characters he gives are suspiciously ju venal 

 in nature, much like those of the type of vinaceigularis, the adult 

 plumage of which was later described as intermedius. 



■>» Nov. Zool., vol. 37, p. 302, 1932. 

 s-Ibis, 1928, p. 264. 



'"Bull. Brit. Orn. Club, vol. 41, p. 18, 1920. 

 "Bull. Brit. Orn. Club, vol. 43, p. 98, 1923. 



