345 
its Applicability to the Ferns, &c. 
exemplified in Engler and Prantl’s ‘ Natiirlichen Pflanzenfamilien has 
been to split up the older families into smaller, and to a certain extent to do 
the same with the genera. This splitting is on the whole fully justifiable, 
and there are yet some cases, it seems to me, where it will have to be done ; 
but, on the other hand, there is no doubt that it may easily be carried too far. 
From my work on age and area it is now fairly clear that in systematic 
work the geographical factor should be taken into consideration to a greater 
extent than has hitherto been the case. In the last twenty years the 
tendency has undoubtedly been in this direction, and the genera and families 
are tending more and more to be split into groups which show geographical 
as well as structural affinities. But it would seem as if more stress yet 
must be laid upon the former. 
The figures given in my various papers on this subject afford no evidence 
to show that any species are actually dying out. Many people at once 
jump to the conclusion that by this I mean that no species are dying out, 
but this is by no means the case. What the figures show is that such cases 
are too few to be seen in them in an unmistakable way — and this fact, by 
the way, shows much more clearly in those sets of figures which (as in the 
case of New Zealand) are the result of actual measurement, than in the 
figures for Ceylon, which were merely estimates. New Zealand may quite 
well contain 20-30 or more species which are in process of dying out, but 
their presence is not shown by the figures, and could not be, unless they all 
belonged to one or two families. 
But if this be so, then it is no longer safe to assume a greater dis- 
continuity in distribution than may be accounted for by known geological 
changes, except in those cases where we may with reasonable probability 
invoke accidental transport by air or water. And no isolated example of 
geographical discontinuity can be accepted as of value without other 
confirmatory cases. Numerous North Temperate Zone species, for example, 
occur again in Australia or New Zealand, without there being any in the 
intermediate regions, so far as we know. This we may perhaps put down to 
alternating glacial periods. Fern spores again, we know, may be carried 
almost indefinite distances by wind, and may germinate if on arrival they 
strike a suitable spot. Probably this does not often happen, but it almost 
certainly does at times occur, so that we may accept as quite possible any 
distribution that may occur for any group of Ferns, without needing any 
other supplementary hypotheses. 
But in the flowering plants such long-distance transport is very rare, 
and we have little evidence to show that it occurs for instance between one 
continent and another. Probably the Orchids and Composites are best 
suited to such transport. On the summit of Ritigala, for example, which is 
separated by about 40 miles of ‘ dry' country from other ‘wet’ mountain 
country in Ceylon, I found (12) 49 wind-carried wet-zone plants. Of these 
