312 
THE KILIM A -NJA B 0 EXPEDITION. 
is different. The Arabs had penetrated very little 
into the hills and valleys of Usambara, so difficult of 
entry in the opposition of a sturdy people, and there¬ 
fore a theoretical cross with them would not satisfac¬ 
torily account for the yellow skins and fine features of 
some of the Wa-sambara. Possibly the aboriginal moun¬ 
taineers, who preceded the Bantu invasion, were allied 
to the Galla or Masai in race, and have largely trans¬ 
mitted their characteristics to the modern inhabitants. 
There are, as it were, two towns of Mazindi: 
one, built on the steep rise of the undercliff, and 
simply the residence and court of Semboja, the 
chief; and the other a regular rambling African 
village in the plain below, surrounded by bauana 
groves and a dense euphorbia hedge. There are be¬ 
sides many gipsy-like encampments of Masai, and even 
temporary “ towns” of those cattle-keeping rovers, 
who have, after years of border warfare, made friends 
with Semboja, and are gradually changing in character 
from restless thievish nomads to honest settled agri¬ 
culturists—in short, here you may see Masai in the 
process of becoming “ Wa-kwavi,” as the Swahili 
traders call them. Far too much has been made of 
this slight distinction. One man has written a voca¬ 
bulary of 66 Kwavi,” and another of “ Masai,” and each 
ignored the fact that they were independently treating 
one identical tongue. Of coarse other grammarians 
and generalizers followed suit like parrots, and so you 
see the Masai family of languages solemnly divided 
into two main branches :— 
(a) Masai, 
( b ) Kwavi; 
which would be as silly as announcing that the English 
“ languages ” offered two distinct groups— 
