208 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



nonvalid and made it a synonym of lo7ir/ipennis, a rather risky pro- 

 ceeding without examining any specimens, but, on the whole, more 

 logical than that followed by Erlanger. Hartert and Goodson ^^ 

 have shown that Grant was mistaken in calling all the birds from 

 Ethiopia 'to East Africa longipennis^ for if they were identical the 

 name dilloni would have to be used for them. However, the latter 

 name really applies to the birds of Eritrea and northern Ethiopia, 

 which are larger than other eastern birds (southern Ethiopia and East 

 Africa). With regard to uhchensis, these two authors feel it wise 

 to provisionally recognize it in the absence of material. They there- 

 fore recognize five races of this pigeon, as follows: guinea (West 

 Africa to Upper Nile) ; longipennis (East Africa to southern Ethio- 

 pia) ; diUoni (northern Ethiopia, Eritrea) \uhehensls (Uhehe district 

 Tanganyika Territory) ; and phaeonota (South Africa). With these 

 conclusions I differ only with regard to longipennis, which I consider 

 indistinguishable from the typical form. The birds of southern 

 Ethiopia are really intermediate between guinea and dilloni, but are 

 closer to the former. I have seen no material of the latter or of 

 uhehensis and suspect that neither are valid races, but do not feel 

 justified in ignoring them without having seen actual specimens. 

 Sclater and Praed o^ " * * * unhesitatingly unite C. g. longipen- 

 nis * * * " with C. g. gumea. Other workers have also found 

 that the so-called eastern form is no good.^^ 



The present series presents the following Aving measurements : 

 Males, 225-243 millimeters; females, 225-23G.5 millimeters (one bird 

 215 millimeters). Hartert and Goodson ^^ give 230-2-17 millimeters 

 for their east African series (one bird 224 millimeters), figures which 

 agree fairly well with mine. They give 230-240 millimeters for west 

 African birds, and nevertheless recognize longipennis (although only 

 as doubtfully separable) in spite of the similarity in measurements. 

 I have seen no western guinea, but only material from the Congo- 

 Uganda border (which Hartert and Goodson state to be typical 

 guinea), but the data all point to the conclusion that it is impossible 

 to separate eastern from western birds. 



A young but fully grov»'n female from Lekiundu River, Kenya 

 Colony, differs from adults in being duller and paler. The neck 

 feathers are not bifurcated; the head is drab instead of clear gray, 

 juost brownish on the crown; the triangular spots on the wing 

 coverts are not pure white but are washed with reddish brown, those 

 of the scapulars and the inner wing coverts being darker than those 

 of the outer coverts; there is also a brownish tinge to the narrow 



•B Nov. ZoOl., vol. 21, 1918, pp. 356-358. 



»«Ibis, 1920, p. 828. 



^ See Zodlitz, Journ. f. Ornith., 1910, pp. 340-341 ; Gyldenstolpe, Kungl. Sv. Vet.-Akad. 

 Handler., 1924, p. 305 ; Van Somoren, Journ. B. Afr. and Uganda Nat. Hist. Soc, No. 30, 

 July, 1927, p. 75, etc. 



