NIGERIA 189 
there, galleries of evergreen vegetation occupying the intersecting 
belts of permanent streams or fringing a marsh, wide meadows bordered 
by low forest, but occupied entirely by grasses of few species, with 
scattered islets of foliage, or, again, outcrops of bare laterite and 
isolated domes and turtle-backs of crystalline rock, ‘“‘ inselbergs ” and 
“ kopjes,” introducing features of their own. 
The Niger Delta displays admirably the high evergreen or moist 
tropical forest. This probably connects itself through the Kameruns with 
the great Central African forest of French and Belgian Congo, which 
is again said to be continuous through the gap between Ruwenzori 
and Lake Albert with that of British East Africa. 
West of the Niger Delta the coast, including Lagos, Dahomey 
and Togo, is sandy, and bush rather than high forest commences almost 
on the shore. Ascending the Niger, one finds that the true “rain 
forest ’’ ends, but not abruptly, in the neighbourhood of Asaba; a 
gradual transition occurs, first to an intermediate type, partly ever- 
green with many large trees. but mingled with those that lose their 
leaves in the months of little rainfall. This change is apparent even 
on the river-bank, and below Lokoja a more open but still semi- 
evergreen forest clothes the valleys, but shows already more of the 
deciduous element on the hills. Farther north the voyage from the 
mouth of the Kaduna River to the Zungeru light railway terminus 
at Barijuko, as often experienced in previous years, reveals again 
the progressively deciduous character of the foliage, resulting in a 
still more open forest. 
The river-bank, however, possessing permanent moisture and 
its own local climate, is an inadequate index of the general features 
over the country at large. To travel by land from Baro to Zungeru, 
and thence either through Kontagora to Sokoto, or by Zaria to Kano 
and Gummel, is to have the complete vision of the West African 
savannah in its various degrees, and, except for local interruptions 
associated with considerations of altitude or geological outcrop, etc., 
or with the lines of perennia] streams, to have it in its regular sequence. 
Leaving out of account for the present the region south of Lokoja 
and of the south bank of the Benué, which in part represents the 
‘* Zone Guinéenne ” of Chevalier, we find that the two routes suggested 
above will take one through country almost entirely of the types 
included by the same writer within the “‘ Zone Soudanienne,” with 
an approach at places along the Anglo-French boundary to the con- 
ditions occurring in his “Zone Sahélienne.” It would be difficult 
and misleading to express these zones definitely in terms of latitude, 
but one might hazard the statement that the Guinea Zone passes to 
the Soudan Zone somewhere between 8° and 10° North, and the latter 
extends either to the northern boundary or verges on a drier belt 
beyond the latitude of 12° or 13° North. If a complete botanical 
