INTRODUCTION. XX VII 
A. furcatum over another, A. horridum in Tahiti, but they seldom 
go together, and are never accompanied by the other related 
members of their group, which in fact have not thus far been 
found anywhere else. Shall we conclude from this that the 
Hawaiian Islands are the original home of A. affine, A. contiguum, 
A. caudatum, etc., and that only the spores of these species have 
been carried by the aid of birds, or whatever other agency, to 
distant countries? This would be little probable and not con- 
sonant with the conclusions we are driven to by the study of 
the other Orders of plants which constitute the Hawaiian flora. 
But if these species as such — ready made — were conveyed 
to the Islands, how does it happen that they fit so exactly in 
the frame of affinity with other species which already existed 
before their arrival in the new home? Perhaps the group of 
A. Kaulfussii will help us in overcoming the difficulty. Here 
we have in the lowest species of the series (and here I comprehend 
for the sake of convenience the form @ of the three first species 
under one head, for they are collateral, all three must be con- 
sidered as parallel first links in the chain of similar development) 
a fern which has great affinity to many other so-called species, 
as is attested by the great number of synonyms it has received 
from pteridologists who saw only one or few species. Its pecu- 
liarities are of apparently trifling importance: thus it differs from 
the New Zealand and Australian A. obtusatum or A. lucidum in 
the character of the scales, the width of the angle at which the 
veins leave the midrib, and the degree of curvature in the latter. 
In the simple form the second and third differences are so slight 
that they barely attract attention. Now it happens that both 
the New Zealand fern and the Hawaiian have parallel lines of 
evolution, a dareoid and a pinnate. But while the simple forms 
are only to be distinguished with difficulty by the slight charac- 
ters indicated above, the higher forms diverge more and more. 
Our dareoid varieties are already sufficiently distinct from A. 
flaccidum, Forst., and our A. Lydgatei and A. meiotomum cannot be 
mistaken for ve difforme, R. Br., A. bulbiferum, Forst., or the 8. 
Polynesian A. multifidum, Br., which all belong to the group of 
A. obtusatum. And then our ay bipinnatum is a fern which has 
no counterpart in that group. In fact, the higher evolutions of 
these two groups, although running parallel, are all unlike each 
other, and only the first links are very sim 
It is otherwise in our groups of 4. peieilalalania and A. 
contiguum. Both are amply represented in the 8. Polynesian 
