BIRDS OF ETHIOPIA AND ICENYA COLONY 323 



Egyptian Sudan and the West Nile Province of northwestern 

 Uganda; erlangeri^ a darker-backed, larger bird, wings 127-135 mm 

 in the male, 122-132 mm in the female, and with the underparts 

 pm'e white with very distinct dark-brown side patches, occurs in 

 northern Somaliland and throughout Ethiopia ; deckeni, the smallest 

 of the races, wings 116-122 mm, similar to erlangeri in color on the 

 underparts, but lighter on the back, is said to live in southern 

 Somaliland and the coastal districts of Kenya Colony south to the 

 mouth of the Tana River ; flscheH^ a race with the dorsal coloration 

 of deckeni^ the ventral color of ruppelU, and larger than either, 

 wings 126-135 mm in the males, 124r-127 mm in the females, in- 

 habits northeastern Tanganyika Territory north to the Ukamba dis- 

 trict of Kenya Colony; while the last form, dohmi, which is like 

 fischeri, but paler above, is known from southwestern Tanganyika 

 Territory. 



Of the so-called forms of rilppeUi, I have examined 56 specimens 

 from the ranges of all five and find it possible to identify and recog- 

 nize only three — rilppeUi, ei'langeri^ and hohtni. Sclater and Mack- 

 worth-Praed and van Someren likewise find deckeni and -fischeri 

 untenable, but they differ in their disposition of these names. Ac- 

 cording to the latter, erlangeri is the race inhabiting all of Ethiopia, 

 Somaliland, Kenya Colony, and northeastern Tanganyika Territory, 

 and of this form deckeni and flscheri are synonyms. Of rilppelli 

 van Someren states that it is smaller than "the other more southern 

 forms", meaning, I presume, deckeni, fischeri, and hohmi. I have 

 examined enough material of ruppelli from near Gondokoro to 

 satisfy myself that it is not smaller than any East African birds, 

 and, on the other hand, I find that Ethiopian birds are uniformly 

 larger than examples from Kenya Colony. Therefore, I consider 

 birds from the latter country to be the same as Gondokoro birds and 

 different from those of Ethiopia. This is essentially the came de- 

 cision as that made by Sclater and Mackworth-Praed, who find, as I 

 do, that — 



* * * the coloration seems to vary considerably with the time of year, 

 and the size is also not a reliable guide. * * * We, therefore consider 

 that E. r. ruppelli ranges from Mongalla through British East Africa, and 

 from Victoria Nyauza to the mouth of the Tana River to the eastern half of 

 German East Africa. On higher gi'ound, as, for instance, near Kilimanjaro 

 and Kenia — the birds have a tendency to be slightly larger and darker. 



The range of erlangeri as given by Zedlitz is correct; that of 

 rilppelli should be extended to include the area he assigns to ftscheH 

 and deskeni and to extreme southern Shoa, while that of hohmi 

 should be extended northeast to Dodoma, whence I have seen two 

 specimens (Loveridge collection). 



