67 
Microlepia is undefinable because generalized. Leucostegia may also be 
primitive, or it is possibly heterogeneous but it is as natural a group as 
is made by combining it with Hurnata, and its union with Davallia seems 
to me still less proper. 
My reasons for believing that the nearest affinity of Oleandra is to 
Hurnata, have already been published. 34 These do not constitute good 
proof, but they are the best evidence we have as to the affinities of 
Oleandra; but Oleandra would appear from its distribution to be the 
older group. 
The mutual affinity of the other genera treated as Davalliece is still 
more dubious. Arthropteris 'seems to me to be very near Lastrcea, in 
which group its first species was described. Our scandent Lastrcea (No. 
1712) shares with Arthropteris the articulate stipe and the terminally 
placed sori. In spite of the very striking resemblance between their 
fronds, comparing, for instance, Arthropteris raniosa with Nephrolepis • 
Lauterbachii or Nephrolepis cor difolia, the affinity of these two genera is 
by no means above doubt, and if one is descended from the other, it is 
not a proven fact that Arthropteris, in spite of its apparently much closer 
affinity to Lastrcea, is the parent, for Nephrolepis is shown by its distribu- 
tion, and still more by its conspicuous morphological isolation, and by 
the diversity of its fructification, to be a very ancient genus. 
N ephrolepis acutifolia is like Schizoloma in two conspicuous characters, 
the articulate pinnae and the unbroken marginal sori, but this is probably 
only a coincidence. The latter genus is an unmistakable relative of 
Lindsay a, and, less intimately, of Odontosoria, but the common ancestry 
of the group is doubtful. Of the three, Odontosoria seems the nearer 
to Microlepia. The group is certainly terrestrial in origin. 
Monachosorwn is, as Diels says, “habituell an Davallia erinnerndes,” 
but the suggestion of Leucostegia is stronger; and this is due almost 
exclusively to their common share in the aspect of Acrophorus and various 
species of Lastrcea ; that is, Monachosorum is more like the generalized 
ferns than like the more highly developed ferns of any tribe, and its 
assignment to any tribe, by our present knowledge, is purely arbitrary. 
The most primitive genus of the Aspleniece is unquestionably Atliy- 
rium. It is a generalized group, sharing, on the one hand, the characters 
of Diplazium and Asplenium, and merging into both, and on the other, 
being indistinguishable from Lastrcea. Athyrium cyclosorum Rupr., of 
Asia and western North America, usually regarded as a form of A. filix- 
fcemina, is Lastraeoid in its indusia. I have recently described a new 
34 Polypodiacese of the Philippines, Govt. Lab. Publ. (1905) 28: 48. “The 
resemblance to * * * the simple species of Hurnata — the creeping, scaly 
rhizome, the articulate stipe, the free, forked, closely parallel veins, the shape, 
attachment, and texture of the indusium, and its opening obliquely toward the 
apex of the frond — all these can not well be construed otherwise than as evidences 
of real affinity.” 
