6 Bews—Some General Principles of Plant Distribution as 
though it is subject to extreme fluctuations, has not altered, materially at 
least, since Upper Cretaceous times. I have pointed out elsewhere how 
tree-veld in South Africa has not, as has often been supposed, replaced 
more mesophytic forest (9), and how our extensive grasslands, though they- 
often give way to forests ( 11 ), yet must have existed for a very long period 
of time, since most of our mammals in South Africa are adapted to grass- 
land conditions (7, p. 150 ). In a word, present-day conditions, in a general 
way, as regards the main types of vegetation, must have existed for a very 
long period of time. 
It is the object of the present paper to consider chiefly present-day 
tendencies in plant distribution in South Africa. The fact that these often 
do throw light on the past will be treated incidentally. Attention, there- 
fore, is directed chiefly to the distribution and origin of species rather than 
to higher groups, a method which differs from that pursued by most 
plant geographers. 
Present-day Conditions in South Africa. 
It is, of course, impossible, within the limits of the present paper, to 
give very full details regarding the distribution of the various types of 
vegetation in South Africa, nor is it necessary for our present purpose. It 
will be sufficient to recall the leading facts. Details have been given by the 
writer in various papers (4-13). There are many distinct climatic areas in 
South Africa, and there is a distinct advantage to be gained by arranging 
them in a series, beginning with the most adverse types and ending with 
the most favourable. Any great subdivision will not be attempted here, 
although, if small differences were taken into account, the series could be 
made a very long one. Even if space permitted, it is doubtful whether 
great subdivision is desirable, for the study of experimental plant ecology 
has shown that environmental differences as regards water content, humidity, 
light, or temperature have to be considerable to cause any distinguishable 
morphological adjustment in plants. They are, as Clements (15) has 
shown, much greater than the unit differences recorded by instruments. 
4 In short, the differences of habitats, as ascertained by thermograph, 
psychrometer, and photometer, are much greater than their efficient differ- 
ences, and, with respect to their ability to produce modification, habitats fall 
into relatively few categories/ In South Africa, too, many species are so 
adaptable as to be able to inhabit rather widely different habitats, e. g. two 
or more of those given below. Though differences in soil conditions are 
efficient in causing changes, yet in South Africa the climatic factors are of 
more importance. The chief climatic areas, roughly arranged in order of 
increasing mesophytism, are the following: 
( 1 ) Western region . The Namib and stony or sandy desert areas of 
the western side. The rainfall is very irregular and small in amount 
