On the Affinities of Smitliornis. 495 



XXVIII. — Some Facts hearing on the Affinities of Smitliornis. 

 By G. L. Bates, M.B.O.U. 



(Plate XVIII. & Text-figure 8.) 



This short paper is written to indicate a few anatomical 

 and other characters of the African genus Smithornis, 

 hitherto placed among the Muscicapidae, which prove that 

 it cannot belong to that family, or, indeed, to the normal 

 Passeres at all. 



If a personal note is not out of place at the beginning, it 

 may be said that the writer has only recently begun, with 

 the help of the ' Vogel'' of Bronn's ' Tierreich ^ and of the 

 articles by Dr. Gadow in Newton^s ^ Dictioiiai'y of Birds,' 

 to make a more thorough study of the birds of this part of 

 Africa (i. e. Kamerun) than that involved in the collection 

 and. identification of specimens. The somewhat scanty 

 observations here recorded are the fruits of that study. 



A brief account of the distribution and habits of the 

 different species of Smithornis will serve to bring the birds 

 in question before the reader. The first species known was 

 S. capensis, which, in spite of its name, is mainly east 

 African, and seems not to be found nearer the Cape than 

 Natal ; but it has been found also in Angola, and recently, 

 by Mr. Neave, in the extreme southern part of Belgian 

 Congo. S. rufolateralis, the smaller of the two forest 

 species, has also been long known, and has a range very 

 extensive in longitude, from Liberia on the west to Lake 

 Albert on the east, but very narrow in latitude. S. sharpei, 

 the largest species, was discovered by Alexander on 

 Fernando Po in 1902, and is figured in ' The Ibis ' for 

 July, 1903. About the same time it was found also in 

 southern Kamerun by Zenker and also by the writer; 

 and a little later, far to the east, by the Ruwenzori 

 Expedition. S. camarunensis, described by Sharpe from the 

 writer's collection in Kamerun and later found by the 

 Ruwenzori Expedition, closely resembles S. capensis, and 

 perhaps should be regarded as a geographical race of that 



2 L 'Z 



