406 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



At first sight this bird would seem to be specifically distinct from 

 minor as the latter has a broad white wing bar and the former has 

 none, but Van Someren ^* has recorded a specimen of cahanisi with 

 the beginning of such a bar. Furthermore, somalicus is interme- 

 diate in this respect and serves to connect them. 



Besides the actual specimens obtained, this bird was recorded as 

 follows: Northern Guaso Nyiro Kiver, July 31 to August 3, 4 

 seen ; Lekiundu River, August 4-8, 22 birds ; Tharaka district, August 

 13, 2 seen; Tana River, August 20-23, 12 noted; west of Ithanga 

 Hills, August 28, 20 seen ; Athi River, August 29, G birds observed. 



Family BUCEROTIDAE 



BYCANISTES CRISTATUS CRISTATUS (Riippell) 



Buceros cristatus Ruppell, N. Wirbelth., Vog., p. 3, pi. 1, 1835: Lake Tsana, 

 Ethiopia. 



Specimens collected: 



One male, Arussi plateau, 8,500 feet, Ethiopia, February 23, 1912. 



One male, Cofali, Ethiopia, March 2, 1912. 



One male, Malke, Ethiopia, March 3, 1912. 



Two females, Aletta, Ethiopia, March 7, 1912. 



One male, Ethiopia, March, 1912. 



The silvery-cheeked hornbill has two geographic races, as follows : 



1. B. c. cristatus. — Ethiopia and the extreme northern parts of 

 Kenya Colony. Although the type locality is Lake Tsana, Sclater ^^ 

 gives southern Ethiopia as the northern limit. 



2. B. c. hrevis. — Central and southern Kenya Colony south through 

 Tanganyika Territory to Nyasaland and Mashonaland. This form 

 differs from the typical one in having a smaller wing length. The 

 two forms meet and intergrade in north-central Kenya Colony. 

 I have seen an intermediate specimen from Nyeri, but birds from 

 Meru and Chuka are true hrevis. 



Grote ^° has suggested that C7^istativs and subcylindricus are repre- 

 sentative members of one form-circle, a statement that is largely' 

 true, but the two forms seem best considered as not conspecific, 

 Grote did not go to the extent of using trinomials, but his map 

 certainly conveys the idea of the geographical representativeness 

 which subspecies ought to have. 



The specimen without locality or definite date is somewhat inter- 

 mediate between cnstatus and hrevis.^ but the wings are in molt and 

 the measurements therefore, not especially significant. The size of 



»^Nov. Zool., vol. 29, 1922, p. 83. 

 ^ Syst. Avium Ethiop., 1924, p. 226. 

 "Mitt. Zool. Mus. Berlin, 1927, p. 205. 



