5 
illustrated by the South African Flora. 
possibility, not merely of the polyphylesis of genera and higher groups, 
which, since the appearance of Engler’s work, has been very generally 
accepted by botanists, but also for species. ‘ De Vries found that Oenothera 
nanella arose from O. lamarckiana , O. laevifolia, and O. scintillans ; 
O. scintillans arose from O. lata and O. lamarckiana ; O. rubrinervis from 
O. lamarckiana , O. laevifolia , O. lata , O. oblonga , O. nanella , and 0. scin- 
tillans , & c.’ Numerous similar examples could bes upplied from the recent 
literature of genetics. 
If the polygenesis of species, particularly, is to be accepted as an estab- 
lished fact, the difficulties of explaining the distribution of many poly- 
genetic species at once disappears ; but, of course, there has been no intention 
on the part of any one of asserting that all species originate in this way. It 
will be shown later that in the case of certain South African endemics 
a polygenetic origin is at least extremely probable. 
Schonland (29) has summarized some of the facts and theories con- 
cerning the origin of the angiospermous flora of South Africa, and he 
remarks at the outset that 4 the question of the origin of our flora having 
been approached almost entirely by European botanists, who were naturally 
influenced by the leading features of the development of the floras of the 
northern hemisphere, it has sometimes greatly suffered through forced 
interpretation of facts and supposed facts (compare, e. g., Thiselton-Dyer, 
1878 )’ (19). Schonland contends that the data for an adequate treatment 
of the subject are absent even to this day, and that there is need for caution 
in dealing with the whole question. Nevertheless, though he kept in mind 
Wallace’s warning that it is 4 so easy and pleasant to speculate on former 
changes of land and sea with which to cut the Gordian knot offered by 
anomalies of distribution ’, Schonland, after considering the facts of distri- 
bution in a number of separate families, found himself forced to the conclu- 
sion that in late Mesozoic times, possibly even up to the Cretaceous 
period, there was still a direct land connexion between Australia and South 
Africa and, possibly even in Tertiary times, between tropical Africa and 
America. 
Recently Schonland (30) has given a summary of the distribution of 
the genera of South African flowering plants with special reference to those 
found in the divisions of Uitenhage and Port Elizabeth without discussing 
further the question of origins. 
The axiom that the present is the key to the past has been accepted 
ever since the time of Lyell, at least theoretically, but biologists have not 
always been willing to use it as much as they should in practice. Not only 
should it be applied to geology and biology, but also to climatology. Until 
we know as much as can be known about present-day tendencies, it is well 
to be cautious about dealing with the past, but I agree with Schonland that, 
so far, all the evidence goes to show that the climate of South Africa, 
