248 Mr. D. A. Bannerman on a CoJlrction of 



Shelley. The addition to the National Collection of 

 Mr. Lowe's two examples from Liberia, which agree exactly 

 with the type of C. accra Shelley, throws considerable light 

 on the subject, and bears out tlie late Mr, Boyd Alexander's 

 theory that C accra and C. fulviventris are perfectly distinct 

 species, and that both may be separated from C. natalensis. 

 In this opinion Dr. Hartert now entirely agrees with me, 

 and I should here like to express my thanks to him for his 

 kindness in placing the material in the Tring Museum at 

 my disposal. 



C. accrcc ditfers from C. fulviventris in the following 

 points, which are mostly alluded to in tlie original descrip- 

 tions of the species. 



C. accrce. 

 (Plate IV. fig. 1.) 



Entire upper parts asliy grey (much 

 greyer than C natalenm'). 



The sides of the face, throat, and 

 breast heavily marlied with dark 

 brown (more pronounced in the 

 two birds from Liberia than in 

 those from the Gold Coast).j 



The abdomen and sides of the 

 the flanlis, which are tawny buif, 

 are heavily barred with brown. 



An ill-detined narrow white malar 

 stripe. 



*Aa in C. natalensis four primaries 

 have a broad white patch. 



The extremities of the primaries 

 are obscurely marked with light 

 greyish-brown. 



The white on the two outer tail- 

 feathers extends .55 mm. from 

 the tip. 



C. futviventris. 

 (Plate IV. fig. 2.) 

 Entire upper parts tawny brown. 



xMuch less heavily marked on the 

 under parts. 



The bars on the abdomen and 

 and flanks are entirely absent. 



Malar stripe broad and pronounced. 



*White wing patch covers five 

 primaries, though less defined 

 on the fifth. 

 No such markings on the primaries. 



Less white on the two outer tail- 

 feathers, extending 43 mm. fi'om 

 the tip. 



* This is a somewhat variable character, and until a large series of 

 C. fulviventris is available cannot be considered of much importance, 

 although the large series of C. natalensis which I have examined is 

 constant in this respect. 



