168 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



been named, none of which appears to be valid. In the present study 

 I have assembled a series of 79 specimens from Ethiopia, various 

 parts of Kenya Colony (Mount Kenya, Escarpment, Naivasha, 

 Chuka, Kericho, Limoru, Fort Hall, Aberdare Mountains, Nyeri, 

 Mau, Gilgil, Lekiundu Kiver, and Mount Garguess), Tanganyika 

 Territory (Kilimanjaro and the Usambara Mountains), and the Ru- 

 wenzori Mountains. I find no constant geographic, size, or color 

 differences that in any way support the contentions of those students 

 who have recognized subspecific groups. Neumann ^^ described salva- 

 dorii from the Kondoa Irangi district of Tanganyika Territory on 

 the basis of their larger size. Three years later ^^ he admitted that 

 the difference was not well marked and that, on the whole, the typical, 

 Ethiopian birds were larger than his race salvadorii. Van Som- 

 cren ®* recognizes salvadorii and also elgonensis Madarasz (supposed 

 to have a wider pectoral band than salvadorii) of Mount Elgon east 

 to Molo and Burnt Forest. He also suggests that a darker race oc- 

 curs in southern Ankole, western Uganda. Granvik,^* on the other 

 hand, relegates Elgon specimens to typical cinnamoineus. Finally, 

 Gyldenstolpe ^'^ examined a good series and could not "detect any 

 colour-differences between specimens from Kilimanjaro, Mount Kenia, 

 Mount Elgon, and the Birunga Volcanoes, nor are there any differ- 

 ences with regard to size." He was not able to examine any Abys- 

 sinian birds and so could not say definitely that there was but one 

 race. I have compared the present three birds from Ethiopia with 

 Kenyan and Tanganyikan material and find no difference between 

 them. Table 33 shows very clearly the absence of any size criterion 

 for the recognition of salvadorii. Only adults are tabulated. 



The color characters are likewise variable; it may be that the 

 degree of rufous increases with age, as Granvik suggested. The two 

 Escarpment birds are darker grayish, less rufous, on the crown and 

 upper back than any others seen, but the difference is a slight one. 



This species is remarkable in that it varies with regard to the 

 number of rectrices, the limits being 10 and 12, thereby bridging the 

 gap between the 10-rectrixed and the 12-rectrixed species of the 

 genus. If not for this species, it might almost be advisable to split 

 the genus into two. Ogilvie-Grant first noticed this and wrote ®® that 

 "examples of this species from the mountains of Shoa, Mount Kenya, 

 and Kikuyu possess 12 tail-feathers, but in a large series of speci- 

 mens from Ruwenzori only 10 tail-feathers are found, though one 



"iJoum. fiir Orn., 1900, p. 304. 



82 Orn. Monatsb., 1903, p. 90-91. 



•=3 Nov. Zool., vol. 25, p. 288, 1918; and vol. 29, p. 230, 1922. 



«Journ. fiir Orn., 1923, Sonderheft, pp. 239-240. 



^Kongl. Svenska Vet.-Akad. Handl., 1924, pp. 137-139. 



96 Trans. Zool. Soc. London, vol. 19, p. 354, 1910. 



