16 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



Zedlitz considers cheniana^ cantillans, and alhicatida all conspecific, 

 but this view can not be maintained. M. clheniana and its races have 

 shorter, more bluntly conical bills than either M. cantiUans or M. 

 alhicauda. The last two can not be anything but distinct species, 

 as they occur together throughout the range of the latter. M. Candida 

 is a fourth species closely allied to these three. It is hard to under- 

 stand how four species of the same subsection of a genus might evolve 

 in the same general region, but we know so little of the habits of 

 any of them that it is impossible even to begin to speculate. 



Erlanger ^ found this lark ver}^ common in the steppes of the 

 Danakil region and also in suitable places in southern Somaliland. 

 It is more a bird of the bush country, less one of the open grasslands, 

 than M. -fischeri Erlanger found the species breeding in May and 

 June. Thus, on May 3, at Karo-Lola in the Garre-Lewin country, 

 southern Gallaland, he found a nest containing four eggs, partly 

 incubated. On June 20 he found a nest with two eggs at Filoa. 

 He describes the eggs as being quite glossy and pale greenish white 

 in ground color, speckled with olive and olive-brown flecks, chiefly 

 around the large pole, and averaging 19 by 15 mm in size. 



Inasmuch as larks have but one complete annual molt a year (the 

 postnuptial one), the fact that the present series, collected in Feb- 

 ruary, are all in worn plumage shows that they were certainly not 

 long through breeding, and may not have even started. 



The four birds from the Hawash River are so much grayer, less 

 brown, above than the adults from Kenya Colony that at first I took 

 them to represent another form, but I have seen similarly grayish 

 birds from the Northern Guaso Nyiro Eiver and from Lake Magadi, 

 in the American Museum of Natural History, and am therefore forced 

 to the conclusion that the difference, startling as it is, is wholly due to 

 wear, the gray birds being much abraded, the brown ones freshly 

 feathered. 



In this connection it may be noted that van Someren '^^ records 

 birds from Lake Magadi as a possible undescribed race, differing 

 from tnarginata in being generally darker, especially on the crown, 

 and in having the dorsal marks less streaky. As mentioned above, I 

 have seen a typical example of marginata from Lake Magadi, and 

 may add that one of the adults from the Indunumara Mountains 

 is so unusually dark above that it looks more like a specimen of 

 M. alhicatida than of M. cantillans marginata. It differs from the 

 former, however, in having brownish edges to the feathers of the 

 dorsum and in having the white on the outer rectrices more restricted, 



8 Journ. fur Orn., 1907, pp. 43-44. 

 "Nov. Zool., vol. 29, p. 178, 1922. 



