614 Transactions of the Royal Society of South Africa. 
branching, general form, shape and size and structure of the leaves, degree 
of succulence and pubescence, all vary according to whether the species 
grows isolated or in the bush, and according to the plant formation of 
which it forms a part — Bush, Stream Bush, Sand-dune Bush, Eocky Hillside, 
or Thorn Veld. A much-branched shrub, given suitable conditions, may 
develop as a small tree. This is what is to be expected, and hardly calls for 
further remark. The tendency of shrubs to become lianes, as already 
mentioned, is more noteworthy. 
(3) The DEciDuors tendency is not perhaps quite so well marked as in 
the case of Mesophanerophytes. The species that show it are deciduous or 
evergreen, according to circumstance's, e.g. Domheya dregeana, Ehretia 
liottentottica, Rhus spp., Excoecaria ofricana, Excoecaria reticulata, Erythrina 
foment osa, Acacia spp. 
(4) As a result, partly perhaps of grass fires, thick bark is a feature of 
most of the species that grow isolated or semi-isolated in the thorn veld, on 
rocky hillsides, or on the outer margin of the bush. 
Another effect of grass fires is seen in Microphanerophytes and also in 
Nanophanerophytes. Many shrubs and even small trees in similar situations 
(veld and outer margin of bush) are constantly burned down to the ground. 
The result is that there is a succession of annual shoots from a permanent 
root crown. 
(5) Influence of Frosts. — On the coast belt, frosts are practically 
unknown, and many species grow there which are not found at the higher 
altitudes inland. The growth-forms of other species differ in the two regions 
(coast and inland) owing to the fact that in the latter they have to endure 
frosts. The upper shoots die off, and the result is as if the shrub had been 
pruned back. New shoots quickly develop in spring, and we thus get a 
succession of annual branches from a certain height. This effect is there- 
fore similar to that of fires, but it is not so drastic. The difference in 
growth-form depending on this factor may be seen not only in comparing 
coast plants with those inland, but also among the inland plants them- 
selves. In the Midlands there are many localities where frosts are rare, 
e.g. sheltered side valleys opening out into a deeper main valley where 
the cold air is drained off. Also the higher ridges and hills, where the 
gradient is sufficient to allow of rapid cold-air drainage, do not suffer so 
much as the main valleys. In such places species are not so much affected 
by frosts. 
(6) Spinosity. — The cjuestion of the cause of thorn development is one 
that has of recent years given rise to a good deal of discussion, and it cannot 
be said, as yet, to have been satisfactorily answered. 
A. Lothelier (1890-93) published the result of researches on Ulex 
euro^aeus, Berheris, Crataegus, etc., in which he showed that dryness of the 
atmosphere and intense illumination favoured thorn production. He found 
