BIRDS OF ETHIOPIA AND KENYA COLONY 241 



Two males and one female, Hawash River, Ethiopia, February 11, 

 1912. 



Five males and four females, Gardula, Ethiopia, March 27-28, 

 1912. 



Two females, Gato River near Gardula, Ethiopia, April 21 to 

 May 11, 1912. 



One male, Bodessa, Ethiopia, May 24, 1912. 



One male, Endoto Mountains, Kenya Colony, July 23, 1912. 



This series contains the types and paratypes of T. chalcospilos 

 intensa (Mearns) and T. chalcospilos Tnedia (Mearns), and one speci- 

 men that Mearns considered as acanthina Oberholser. The present 

 species has been divided into as many as nine different races, eight of 

 which are apparently untenable. The Damaraland form, volhmanni^ 

 which I have not seen, is regarded as distinct by Sclater, who has 

 studied these races more recently and with better material than any- 

 one else. Besides the above-listed specimens, I have examined 15 

 others, making 35 in all, an admittedly inadequate series. However, 

 even within this series, individual, nongeographic variation is great 

 enough to cast serious doubt on the validity of such forms as 

 acanthina^ inedia^ intensa^ and sovial'ica. 



Some of the forms were described on erroneous assumptions. For 

 instance, Oberholser ^^ described acanthina on the assumption that 

 west African birds were typical chalcospilos. He was of the opinion 

 that Wagler's description was based on " La Tourterelle du Senegal " 

 of Brisson.^"^ However, more recently Sclater" carefully examined 

 Wagler's original description and found that he " * * * un- 

 doubtedly based his description chiefly on Levaillant's plate, and 

 that the type locality should be South Africa, as he states in his last 

 paragraph : ' Habitat in Africa meridionali satis f requens in terra 

 Caffrorum porro in Senegambia ' * * *." Therefore, it follows 

 that the eastern Cape Province (Caffraria) should be the type local- 

 ity. Furthermore, the green-spotted wood dove is not known to 

 occur in any part of West Africa except Angola and Loango, so 

 Senegal could hardly be the type locality of chalcospilos. 



Reichenow's name ca-ffr^a therefore becomes a direct synonym of 

 chalcospilos, having the same type locality. 



Erlanger's form soTrwIica described ^^ from Salakle, southern 

 Somaliland, seems a very well marked form if the colored i)late only 

 be examined. However, a specimen from British Somaliland is 

 certainly not in any way different from others from Ethiopia, Kenya 

 Colony, and East Africa generally. 



- Proc. U. S. Nat. Mu«., vol. 28, 1905, p. 8i5. 

 ^Ornith., vol. 1, 1760, p. 122, pi. 10, flg. 1. 

 ■"Bull. Brit. Orn. CI., vol. 42, 1022, p. 118. 

 'sjourn. f. Ornitli., 1905, p. i:'.4. 



