274f Letters, Extracts, and Notes. [Ibis, 



as far south as Usakos, which is nearly the southern limit of 

 the busli country. I have no specimens from this place, 

 but when there I saw numbers which appeared very white 

 on the head and neck. 



The most southerly specimen I have is one from Omaruru, 

 a male ; this differs from coffer in having the whole sides of 

 head and neck pure white, with only a very slight admixture 

 of grey feathers on the sides of the neck, in having a much 

 lai'ger amount of white on tlie secondaries and wing-coverts, 

 and in having the outer vanes of the two outer tail-feathers 

 pure white ; the bill also appears shorter and thicker. 

 Yours truly, 

 Tsumeb, C. J. Finch-Davies, Lieut., 



S."\Y. African Protectorate. 1st S A M R. 



o Jauuarv, 1U17. 



The Steamer-Duck. 



Sill, — May I be allowed to write a few lines in *The Ibis* 

 in reply to Mr. Phillips's letter about the Steamer-Duck 

 published in the January number and in which my name is 

 mentioned? It seems evident that Mr. Phillips's collector, 

 Mr. Brooks, has only seen one species of Steamer-Duck and 

 that Tachyeres patachonicus. 



Mr. Phillips has therefore experience of one species only, 

 and accordingly it may perhaps be doubted if he can 

 reasonably talk with authority about two ! Mr. Phillips's 

 collector has only been to tlie Falkland Islands, which are a 

 stronghold of Tachyeres patachonicus. This is confirmed 

 by everything he says about tlie sulyect. Soutli America, 

 however, is a big place, and it is unsafe to ignore the fact 

 that there are other places than the Falklands where other 

 species of Steamer-Ducks may occur. I refer to my papers 

 in 'The Ibis' for July 1916 on the subject and wish to 

 confirm every word I have written there. 



That Mr. Brooks has not seen T. jjatachonicus fly and 

 could not make them fly although he has seen " thousands," 

 notwithstaxiding that those birds were, judging from his 

 description, really T. patachotiicus, only tends to prove that 



