350 BULLETIiSr 15 3, UNITED STATES NATIONAL MUSEUM 



Edward, and South Victoria Nyaiiza." On the following page he 

 lists specimens from Kendu Bay south of Speke's Gulf, in Tan- 

 ganyika Territory, and from South Ankole in Uganda as " H. 

 'pallidivcntri8{'l) subsp. nov.," and then goes on to say that these 

 birds belong," * * * to the smaller, pale-bellied, violet-winged 

 birds of the central lake district which I have named above * * *. 

 The birds collected by the RuAvenzori expedition belong to this race." 

 This is all there is by way of a description ; no type or type locality 

 is mentioned, and Hartert does not include the type of kivuensh 

 in his list of the types of birds in the Tring Museum. A series of 17 

 specimens of 'pallidiventris from points as widely separated as 

 Kilosa, Tanganyika Territory, and Luluabourg on the Kasai show 

 no differences, and it therefore appears highly improbable that 

 kivuensis is valid. 



Sclater,^'' in defense of his action in considering paUJdiventris 

 a race of leucocephala, writes that it occurs south of the Zambesi 

 only in the rainy season (October to March), and in the southern 

 winter (April to September) to the Semliki River and the Ruwen- 

 zori Mountains. In this he is mistaken as I have seen undoubted 

 examples of pallidiventris from tropical East and Central Africa 

 taken in every month except February. Furthermore, a female taken 

 on August 3 at Medje, northern Belgian Congo, by J. P. Chapin had 

 the ovary slightly enlarged. (However, Chapin informs me that 

 'pallidiventris does not breed north of the Congo forest.) 



This species ranges farther north than most writers state. In 

 Kenya Colony (from which country it does not appear to have been 

 previously recorded) it is known from Kisumu and Kenna Tana 

 (specimens in Museum of Comparative Zoology), and in the Belgian 

 Congo as far north as Medje. 



Van Someren considers ogUviei a race of pallidiventris and Sclater 

 treats it as a synonym of the latter. However, in the original de- 

 scription it is said to differ from leucocephala in having the wings 

 and tail almost ])ure violet as in pallidiventris^ and from the latter 

 in having the belly dark deep chestnut as in leucocephala. If the 

 color of the wings and tail is to be taken as the specific character 

 then both Van Someren and Sclater are inconsistent in considering 

 hyacinthina a form of leucocephala. Furthermore, if we use the 

 remigial and rectricial color as a specific criterion we are left with the 

 problem of dark-bellied and light-bellied birds occupying the same 

 area. It seems therefore that oqilviei is to be considered the same 

 as hyacinthina rather than pallidiventris. 



2. Halcyon leucocephala leucocephala. — I can see no great ad- 

 vantage in separating Senegalese from Ethiopian and Kenian birds. 

 However, I have seen but one specimen from Senegal and only one 



*» Syst. Avium Ethiop., 1924, pp. 216-217. 



