100 BULLETIN" 15 3, UNITED STATES NATIONAL MUSEUM 



Male, Ourso, Ethiopia, October 15, 1911. 



Three females and two males, Dire Daoua, Ethiopia, November 

 2 to December 17, 1911. 



Male and female, Guaso Nyiro River, Kenya Colony, August 1-2, 

 1912. 



The colors of the soft parts are recorded for an adult female 

 (from Guaso Nyiro River) as follows: Feet orange; claws black; 

 cere and orbit vinous red; bill blue with black tip. 



Our knowledge of the races of the African pygmy falcon is rather 

 unsatisfactory in that while most authorities agree in recognizing but 

 two forms — seimtorquatus and castanonotus^ the latter seems but 

 slightly more valid than deckeni, which is usually regarded as un- 

 recognizable. It may be that deckeni seems more valid to me than 

 to other writers because I have seen but three specimens from Soma- 

 liland, all of which happen to fit the description. However, I feel 

 that the species PoliMerax semitorquatus does not vary enough geo- 

 graphically to warrant the recognition of many local forms. Ac- 

 cording to Swann ^^ and Sclater,"^ castanonotus ranges south to Kenya 

 Colony, Uganda, and Tanganyika Territory, while semitorquatus 

 does not occur north of Damaraland, Bechuanaland, the Orange Free 

 State, and Basutoland. On tlie other hand, Van Someren ~^' writes 

 that seTnitorquatus ranges north to Kenya Colony (Athi River). 



In the Museum of Comparative Zoology there is a specimen, un- 

 doubtedly of the typical race, from Mwanza, Tanganyika Territory, 

 indicating that the characters of that race may appear far to the 

 north in equatorial Africa. This specimen is large even for typical 

 semitorquatus, and is in no sense an intermediate. 



The race which Oberholser described and named homopterus is 

 recognized, among recent writers, only by Zedlitz. On geographic 

 grounds the Abyssinian specimens in the present collection should 

 belong to this race, if it were valid. Zedlitz records a male speci- 

 men ^^ from Dire Daoua with a wing length of 122 millimeters (a 

 slight excess over either of the two birds in the Frick collection), 

 and eight other males from Ethiopia with wings of from 116-123 

 millimeters. 



These limits of size variation lie wholly within those of castano- 

 notus (in which the wing varies from llG-124 millimeters). On 

 the other hand. Van Someren -^ gives the wing variation of his series 

 of male homopterus^ as 110-115 millimeters. In this case the limits 

 are included in those given by Zedlitz for his race decheni. The 



=* Synopsis of the Acclpltres, 1922, p. 1S3. 

 2s Syst. Avium Ethiop., 1924, p. 56. 

 =« Nov. Zoo]., vol. 29, 1922, p. 44. 

 =^Journ. fiir Ornitli., 1914, p. 676. 

 2" Nov. Zool., vol. 29, 1922, p. 44. 



