64 
oceasioual in Lasiraa, are common in Gonx opt oris, and liair-like triehomos 
are found in some species on frond, indnsiiim, and even on the sporangia. 
The groiip is also characterized on the whole by numerous small stoniata. 
Mesochlaena is a Goniopteris in form, texture and even in the charac- 
teristic pubescence, and, in a genus notable for the instahility of indumim 
characters should hardly be separated by the indusium alone. As a 
matter of fact, the more fundamental peculiarity is the elongation of the 
sorus, which is very moderate, but sufficient on mechanical grounds to 
effect the division of the indusium as in Diplazium. The recognition of 
Meniscium as a genus characterized by elongate sori is generally and 
properly abandoned. Though a very large and pantropic group, Goniop- 
teris has a narrower range than Lastrcta, 
Both Polystichum and (through Pleocnemia?) Aspidium are so 
intimately related to Lastra>a that the proper assignment of species is 
sometimes difficult and able botanists are not wanting who would still 
include all in one great genus. ■ Pohjstichum is probably older ^ than As- 
pidium, being cosmopolitan, while Aspidium is tropical, and rather nearer 
i 
in character to Lastrcea, but the indusium characters of Pohjstichum 
are hardly as unstable as those of Aspidium. In the latter the reticulate 
•venation is directly correlated with the simplification of the frond. The 
indusium furnishes valid characters for most species of Aspidium, and, 
like any other character, is to some extent a guide to affinities; but the 
distinction of Sagenia, Tectaria and Arcijpteris^ genera to be recognized 
by this character alone, breaks up some certainly natural groups. I 
am not ready to judge the real naturalness of Pleocnemio., as a genus. 
Judging by its distribution, Didymochlaena moist be a rather old 
genus. Its aspect suggests that it may be descended through Pohjstichum. 
CyclopeltiSy ia likewise homogeneous genus in the tropics of both hemi- 
spheres, is generally recognized as an offshoot of Polysticlium. 
In the group of small genera split but of Achrostichum and included 
by Diels in Aspidiew, dimorpliism has so obscured most other characters 
that it is not easy to be sure of their real affinity. However, the group 
as a whole is almost certainly unnatural, for I can not imagine our two 
so-called Polyhotryw to be congeneric, or that both are intimately related 
to Stenosemia or Leptochihis. In view of their conspicuous morpho% 
logical and geographical isolation, both of the former oiiglit clearly to 
be restored to generic rank, as Egenolfia and Psomiocarpa. The former 
nui\-, with a high measure of probability, be regarded as derived from 
Polystichnnt, in which genus P. auriculatum is the most similar species 
in this part of the world; but a direct Lastrcea ancestry is not impossible 
and it sometimes approaches Leptochilus. I^either tlie loss of the 
indusium, nor a considerable dimorphism is any novelty in Pohjstichum. 
Psomiocarpa is possibly more doubtful, but I do not believe that 
*°Tliat is, as a group with its preijent characters; all plants may be assumed to 
have an equally long ancestry. 
\ 
