illustrated by the Soitth African Flora. 1 1 
by birds. In other cases, where there is present the widely distributed and 
closely related species, polygenesis would appear to afford a much simpler 
explanation. It would appear, therefore, that we must admit at least the 
possibility of polygenesis, for, if it affords a simpler explanation than 
ordinary migration over enormous areas, then the onus of proof is on the 
side of those who deny the possibility of multiple origins. Personally, 
I would go farther and advocate its extreme probability. As we have seen, 
Clements regards it as definitely proved by the work of de Vries. 
If now we extend the facts learned about the species to the genus, 
tribe, and family, we may take it for granted that the same methods will 
apply. We find widespread families like the Scrophulariaceae have given 
rise to families with a restricted distribution like the Selaginaceae, which is 
mostly S. African, with a few isolated genera to the north, or the Myopo- 
raceae, which are mostly Australian, with isolated examples elsewhere as far 
apart as China and the West Indies, and one genus Oftia in S. Africa. 
Some authorities sink these families as tribes of the Scrophulariaceae, but 
this does not affect our argument. Again, in the Verbenaceae, Iridaceae, 
and other families we find the oldest tribes or sections of the family con- 
centrated in the south-western region of South Africa, while other younger 
tribes or sections are eastern. All these facts are very significant. Instead 
of beginning with the whole world’s flora and finding connexions between 
the Cape and Australia, for instance, we begin with a restricted area and 
study the origin of species. Then with caution our methods and observa- 
tions may be extended to a wider area and the time lengthened from the 
present into the past, always keeping in mind that the only key to the past 
is to be found in the present. 
The Evidence. 
It may be well to anticipate one or two obvious objections which may 
be brought forward to accepting certain of the following examples as 
evidence supporting the above conclusions. The first is that our knowledge 
of detailed distribution in South Africa is still far from complete. Species 
which are now thought to be restricted or discontinuous in their distribution 
may very well subsequently be shown not to be so. This objection is valid, 
no doubt, in some cases, but it certainly cannot apply to them all. Its weight 
as an argument is in inverse proportion to the number of examples given. 
The second objection is that the distribution in each example given can 
be explained by ordinary migration and its abundance or rarity by adapta- 
tion to environmental conditions. As regards the first part of the objection, 
it is, as already admitted, difficult to disprove in any particular case, nor do 
I seek to deny that it is usually a sufficient explanation, especially in the 
case of species distributed by birds. As regards the second part of the 
