Geography and Ethnology, 261 
the hills, in which the Sabaki has it sources, and that 
the Tsavo, Chavo, and Adi, which really do derive their 
w^aters wholly, or in part, from Kilima Njaro, run into 
or form the Ozi. Indeed, the word Adi appears to us 
to be identical with Ozi, and to differ from it only 
through that dialectical peculiarity of pronunciation 
which has made Dr. Krapf write Dana also for Tana, 
the Pokomo word for that river." * 
This, with all respect for Mr. Cooley s character as 
a geographer, is a perfect jumble of mistakes. First, 
the Sabaki does enter the " Bay of Malinde," and the 
waters of this bay are often discoloured by the Sabaki's 
floods. Then why, if an obsolete," should the Bay 
of MaHnde " be an incorrect expression"? Malinde 
is situated in a bay, how then can it be incorrect to 
call that bay the Bay of Malinde ".^^ Secondly, if 
the Sabaki does enter the Bay of Malinde,'' its true 
position cannot be at Pamamba. The Pamamba is an 
insignificant creek, or Mukono wa Bahari (an arm of 
the sea), which extends only a few miles into the ~ 
country, and over the end of which we passed, or 
rather stepped, on our way to the Tana. Thirdly, 
Pamamba is not the correct designation of the place ; 
it is called Pa-M^amba, Moreover *'mamba" is the 
Kisuahili word for crocodile, not hippopotamus ; but 
muamba" signifies rock. Pa-Muamba then means a 
place of rocks, and not, as Mr, C. affirms, hippopo- 
tamus place. Fourthly, if there was formerly a town of 
Sabaki, which ''stood on a hill visible to the north-west 
from Malinde," there is no such town visible there now. 
If there was such a town at any time, I am not pre- 
pared to say that the river might not have taken the 
Cooley^s Inner Africa laid open," p, 1 1 5, 
