348 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



reached and followed by various workers. The reasons for this 

 great diversity of opinion are two — first, that two distinct but related 

 species were for a long time considered conspecific, and second, that 

 both are variable. The resulting difficulty of studying two variable 

 entities erroneously considered as a single specific aggregate will 

 readily be apparent. Neumann ^^ recognized four forms — semi- 

 caerulea of West Africa, northeast Africa, and Arabia, centralis of 

 Masailand, hyacinthina of Zanzibar, southern Tanganyika Territory 

 and Nyasaland, and swainsoni of Angola, South Africa, and East 

 Africa as far north as the interior of Nyasaland and Uhehe. 

 Erlanger "^ separated Senegalese birds from Abyssinian ones and 

 called the former by Swainson's name 7'ufiventris, the latter semi- 

 caerulea. He makes no mention, direct or indirect, of centralis^ and 

 questions the validity of hyacinthina. Reichenow^" first recog- 

 nized swainsoni as specifically distinct, but he admitted three forms 

 of semicaemlea — the typical northeast African one, which, accord- 

 ing to him extends south to the Pangani River, the West African 

 rufiventrls., *and violet- winged hyacinthina. Zedlitz ^^ combined 

 rufiventris with semicaemlea and considered centralis a synonym of 

 hyacinthina., giving to the latter the range of "East Africa," and 

 recognized swainsoni as a southern subspecies of sonicaerulea. 



In 1915 Claude Grant ^^ reviewed the group with the aid of the 

 material in the British Museum and decided that there were six 

 valid races. He restricted the name semicaerulea to the Arabian 

 birds and resurrected the older name leucocej)hala for the whole 

 species. Both centralis and ruflventris he disposed of as synonyms of 

 leucocepJuda; he described ^^ a new race ogilviei from Angoniland 

 intermediate in character between leucoccphala and sioaiiisoni', and 

 recognized swainsoni of southern Africa and acteon of the Cape 

 Verde Islands. With the latter he synonymized erythrorhynchus 

 Gould as had been done by Pucheran and Hnrtlaub some GO years 

 before. Hyacinthina he retained but said that he had seen no mate- 

 rial. Seven years later Van Someren "* showed that the name 

 swainsoni was not applicable to the pale-bellied southern birds but 

 was a synonym of leucocephala. As a matter of fact, Ogilvie-Grant ^^ 

 first pointed this out, but Claude Grant apparently overlooked it. 

 Van Someren revived the name pallidiventris for the pale-bellied 

 birds and showed that they were specifically distinct from leucoce- 



'«.Tourn. f. Ornith., 1905, p. 190. 



™Idem, p. 44C. 



soVOg. Afr., vol. 2, pp. 276-279. 



«i Journ. f. Ornith., 1910, p. 766. 



«2 Ibis, pp. 205-267. 



83 Bull. Brit. Orn. CI., vol. 35, 1914, p. 28. 



ft»Nov. Zool., vol. 29, 1922, pp. 76-78. 



» Trans. Zool. Soc. Lend., 1910, p. 437. 



