214 TRANSACTIONS, NATURAL HISTORY SOCIETY OF GLASGOW, 
Tt follows, then, that no common dividing lines are to be 
expected, and that, in those cases at least which certainly make 
up 12 per cent., and may form 80 per cent., of the British Flora, 
river-valleys exercise no influence whatever. So far, I have 
simply been treating of the British Flora, but during my botanical 
journeys I have myself seen many dividing lines of the floras in a 
larger sense. In South Africa there are at least four; in West 
Africa there are two very distinct ones, the upland or Upper 
Niger Flora seen at Falaba, and the Coast Flora ; in Egypt there 
are distinctly three; in Madagascar three; in Central Africa, 
from Mombassa to Uganda, and thence by Tanganyika to the 
Zambesi, seven different floras, more distinct in each case than 
the floras of Italy and Scotland.* 
In every single case the dividing lines are indistinct; they 
remind one of the very excellent description by Mr. Miller Christy 
of the range of Primula elatior, Jacq., in Britain.* 
The flora, as a whole, is defined by certain general climatic 
considerations, usually very obvious if one takes the amount 
and distribution in time of the rainfall, but in only one single 
case did I feel that the line could be drawn on the ground. That 
was in Egypt, where the boundary of the Nile overflow was 
clearly marked by the vegetation. 
In all the other cases there is a debatable land; the rainfall is 
not the same every year, but varies yearly within pretty broad 
limits, and, no doubt, is also slowly changing over a series of 
years in one or the other direction. 
Let me take two examples from my own experience—one well 
known, and one never, to my knowledge, published. Mr. Bolus 
has thoroughly described the Cape Flora.* The characteristic 
Cliffortias, Ericas, Proteas, &c., of the Cape Peninsula are not 
found in the Karoo, because the Karoo has no rain, and at a 
certain season the Cape Peninsula has. Yet there is a broad 
debatable land in which the two floras mingle, quite impossible 
to mark definitely, and with, I think, a few endemic species not 
found either in the Karoo or in the Cape Peninsula, The second 

1] have not included special Alpine belts. 
2 Journ. Linn. Soc., Vol. XXXIII., p. 172. 
3 Handbook to the Cape Colony. Also Trans. and Proc. Bot. Soc. Edin. 
Paper read November 14th, 1889, 
