64 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I46 



specimens, Rand, Friedmann, and Traylor (1959, p. 271) noted the 

 great variation in the underwing coverts and axillars; from almost 

 wholly dull white to largely blackish. The Tembura specimen in 

 London had no blackish at all under the wings. Years ago, a specimen 

 from Danger River, Gabon, was used as the basis for the description 

 of a very pallid "race" caroli. This specimen was studied in 1962 

 together with the Tembura one, as well as with a very extensive series 

 of others. It was found to be paler, less streaked with black on the 

 chin and throat, but the difference between it and the Tembura bird 

 was not great enough to suggest that it might represent a distinct 

 race. The type of caroli had the terminal white spots on the rectrices 

 larger than in the Tembura specimen. The extreme pallor of caroli 

 suggests, if anything, that just as in coastal Kenya levaillantii may 

 produce completely melanistic phases (albonofatus) , so elsewhere 

 it may almost approach jacohinus in its lack of dark ventral mark- 

 ings. Certainly the geographic distribution of color extremes — 

 darkest birds from Ethiopia, Somali Republic, Kenya, Sudan, Tan- 

 ganyika, and Damaraland in South-West Africa, and lightest ones 

 from Sudan, Gabon, Portuguese Guinea, and Rhodesia — indicate that 

 they are haphazard in their occurrence, and hence not significant 

 taxonomically. In none of these areas are the birds uniform in their 

 variational trends. 



Another variable character of the "normal" phase of levaillantii 

 is the length of the feathers forming the crest. Here again, examina- 

 tion of long series from all parts of the range, tends to rule out the 

 supposed significance of any local extremes. At one time in these 

 investigations it seemed that birds from Ethiopia tended to have on 

 the average longer crests than birds from elsewhere, but measure- 

 ments failed to corroborate this. 



The melanistic phase, originally described as a separate species 

 under the name C. alhonotatus, has been found (with one exception) 

 only in the narrow coastal belt of Kenya, south to Usambara Hills, in 

 Tanganyika, and north to southern Jubaland in the Somali Republic. 

 I have been able to examine 21 of the 26 recorded specimens known 

 to me. Since relatively few investigators have examined this plumage, 

 and none with as ample material, the following notes on its variations 

 are here recorded. 



In general this phase may be described as being black all over, 

 except for terminal white spots on the outer rectrices (this varies 

 from the two outermost pairs to the four outermost pairs in different 

 individuals), and a white patch on the inner webs of the eight outer 



