46 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



agree that senegalensis and melanocrissa were only racially distinct. 

 The truth of the matter is that specific characters in swallows are 

 generally finer and smaller than in many other birds, and there is 

 nothing to be gained by attempting to reduce species to races merely 

 to conform with other groups. 



A study of the characters by which melanocrissa differs from emini 

 supports the disposition of these two made above. The former is 

 said to be noticeably lighter below and has dusky shaft streaks on 

 the feathers of the throat, breast, and upper abdomen. As already 

 noted, however, the shaft streaks are not always present in melano- 

 crissa, and I have seen some dark-bellied emini with faint dusky 

 shaft streaks. Van Someren ^® writes that a female emini from 

 Kenya Colony has the throat and breast streaked, so apparently the 

 birds I have seen are not unusual. 



The only difference that appears to be constant is the color of the 

 underparts. Neumann " records that the color of the upperparts is 

 steel blue in emini and purplish blue in melanocrissa. This does not 

 hold true in the series I have examined (20 birds in all), and Gylden- 

 stolpe likewise was unable to corroborate Neumann's observation. 



The range of melanocrissa is as follows: Northern and central 

 Ethiopia. As far as I know there are no definite records from 

 Eritrea, but Brehm observed a swallow that he considered to be 

 melanocrissa at Mensa, Bogosland. Von Heuglin, however, cast 

 doubt on this identification and suggested that the species observed 

 might have been Hirundo senegalensis. Besides the numerous rec- 

 ords from Ethiopia, I have seen one bird from Juja Farm, Athi 

 River, Kenya Colony, that is typical melanocrissa. This is not only 

 the southernmost record for the species and the first one for Kenya 

 Colony, but is probably to be interpreted as an accidental one, i. e., 

 the species is certainly not usually found so far south. This bird 

 was collected by Mearns while he was with the Smithsonian African 

 expedition under Colonel Roosevelt, on May 13, 1909. 



Typical rufida occurs in the Egyptian Sudan, Eritrea, and the 

 Yemen Province of Arabia, and domicella is found in the southern 

 Anglo-Egyptian Sudan (Mongalla, White Nile, Bahr el Ghazal, 

 etc.) west through Darfur to the savannahs of Upper Guinea. The 

 range of emini will be dealt with under that form. 



The published notes on the ecological habitat and habits of this 

 swallow are not sufficiently harmonious to enable us to get a very 

 clear picture. Riippell found the species on the high plateau of 

 Temben and in the Simien district. Von Heuglin found it in cen- 

 tral Ethiopia throughout the rainy season until February, both in the 

 mountains and plains, and noted that it seemed to depart between 



«Nov. Zool., vol. 29, p. 92, 1922. 

 "Journ. fur Orn., 1905, p. 202. 



