458 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



parative Zoology, the American Museum of Natural History, and 

 the Field Museum, and the conclusions arrived at completely sub- 

 stantiate and corroborate Neumann's work. The only change to be 

 made is the northern limit of the range of jacksonl. Neumann states 

 that this form occurs north to the Kenya-Ethiopia boundary, while 

 the series collected by Mearns at or near Gardula, Bodessa, Tertale, 

 and Mar Mora, extend the range into southern Shoa. They are the 

 first specimens of jacksonl recorded from Ethiopia, and are not in- 

 termediate in nature between typical jacksonl and gallai'um, but are 

 practically identical with north Kenian material, slightly approach- 

 ing versicolor^ but agree fairly well with a topotype of jacksoni. 

 The races recognized have the following ranges and characters. 



1. T. e. eri/throce'phahis. — Northern Tanganyika Territory from 

 the region of Lakes Manyara and Nguruman to the Kilimanjaro 

 region, north to the Taru Desert, the Teita, Ukaniba, and southern 

 Kikuyu districts, Kenya Colony. Lake Naivasha is the westernmost 

 locality from which I have seen typical erythrocephalv^. This is the 

 largest of all the races, wings 96-100 millimeters, and has a well de- 

 fined black throat patch in the male, and the red of the cheeks and 

 sides of the head continues around the hind edge of the occiput, 

 completely surrounding the black crown patch in the males; the 

 feathers of the crown in the females with small black tips. 



2. T. e. versicolor. — Kenya Colony from Lakes Baringo and Han- 

 nington, north through Barsaloi to the south end of Lake Rudolf, 

 west through eastern and northern Uganda to the country east of 

 Gondokoro on the upper White Nile, and to Turkanalancl. Some- 

 what smaller than the typical race, wings 91-96 millimeters, and with 

 a great deal of yellow on the sides, top, and back of the head (also 

 the superciliaries and forehead), but with the red completely sur- 

 rounding the black crown patch, and with the under tail coverts yel- 

 low or orange red much mixed with yellow. Grant " and Van 

 Someren ®^ write that versicolor is not a valid form, and Sclater ^* 

 follows their lead and rejects it as a synonym of erythrocephalus. 

 However, as Neumann has shown, the form is perfectly valid, and I 

 find it recognizable in specimens (4) from Lake Hannington and 

 Barsaloi. 



3. T. e. jacksonl. — From the Sabaki and Tana Rivers northwest 

 through western Jubaland to the Rendili country east of Lake 

 Rudolf, north to Gardula, in southern Shoa, wdiere it intergrades 

 with versicolor. This form is smaller than the nominate race, wings 

 89-96 millimeters (Neumann states 86-89 millimeters, but of 23 speci- 



52 Ibis, Ifll.^, p. US. 



53 Nov. Zool., vol. 29, 1922, p. 60. 



" Syst. Avium Ethiop., 1924. p. 285. 



