1918.] the Birds of the Anglo-Egyptian Su(Ia?i. 4.67 



With regard to Serinus madaraszi Reichw. O. M. 1902, 

 p. 8, from north of Lake N3'asa, we have do specimens 

 which in any way approacli his description, all the Nyasa- 

 land birds before us beinj^ indistinguishable from those 

 from South Africa. From its description, however, it 

 would seem quite a distinct form. 



The birds from St. Thomas Island which were identified 

 by Bannerman (Ibis, 1915, p. 101) as Serinus hartlauhi, have 

 no sign of grey on the head and do not exactly agree with 

 any mainland race. 



Poliospiza leucopygia leucopygia. 



Serinus leucopi/gius (Sundev.) ; Shelley, B. A. iii. p. 216. 



[B. coll.] 1 Shaslieina nr. Gedaref May, Kas, ; 2 Roseires 

 Aug. Sen. ; 2 Mongalla. 



[C. & L. coll.] 5 Kamisa Dec. Sen. ; 1 nr. Jebelein Jan. 

 W.N. 



The birds from Mongalla are whiter and paler than those 

 from Roseires and approach P. /. rigyenbachi Neumann (Bull. 

 B. O. C. xxi. 1908, p. 4^ : Dakar). 



There would appear to be no structural diftercnces which 

 can be used to separate Poliospiza from Serinus. We propose 

 therefore to follow Reiclienow, and to call all birds which are 

 green above and yellow below Serinus and the rest Poliospiza. 



Poliospiza gularis elgonensis. 



Poliospiza elgontnsis O. -Grant, Bull. B. O. C. xxxi. 1912, 

 p. 17: Mt. Elgon. 



Poliospiza canicapilla Butler (nee Du Bus), Ibis, 1908, 

 p. 218. 



[B. coll.] 1 Br.ingi's Jan. E.G. 



[Chr. coll.] 4 Yei Nov. & Dec. L.E. ; 1 Yambio Mch. 

 E.G. 



In working out the races of P. gularis, we found that one 

 or two other species had at difi'erent times been confused 

 M'ith it. These were Poliospiza tristriata Riipp. from 

 Somaliland and Ai)yssinia, Mhich can always be distin- 

 guished by its smaller size, wing 65 mm. against 75 or over, 

 and the fact that no white shows through on the crown; 



