INTRODUCTION TO 
294 
familiar with the groups may identify the members of 
different groups by convenient characters which do not 
serve for the recognition of groups as a whole. 
The third difficulty is our own ignorance. We 
outgrow this, year by year, and are able, with each year’s 
progress, to make our classification a little more natural. 
The recognition of Taenitis and Ptevidium , as descendants 
(along very distinct lines) of Denustaedtia, is a result of 
quite recent progress. The classification presented in 
this paper is decidedly more natural, l believe, than any 
general one previously used; and yet, even the largest of 
all the fern genera, as construed here, Polypodium, i$ 
located by convenience rather than by understanding; I 
do not feel sure even as to whether it should be regarded 
as more nearly related to the descendants of the 
Matoniaceae or of Denustaedtia. Nephrolepis has been 
no less of a puzzle; but the work I have just been doing 
on Bornean ferns has shown its general line of descent to 
my fairly complete satisfaction. 
Borneo occupies the centre of the world’s greatest 
fern area. Of all parts of the earth, it is the one whose 
living ferns are most likely to reveal the relationship of 
groups and to perfect our knowledge of their descent. If 
these keys encourage fern work in Borneo, and result in 
increased activity in the collection and study of Borneo 
ferns, their preparation will have been much more than 
worth while. 
I have added an Appendix for the assistance of 
those who are perhaps unfamiliar with all the technical 
terms used in the following pages. 
