71 
r 
believe this of one rather tlian of another, and all of them known to 
me impress me as being of almost certain Polypodioid derivation. P. 
Mcrritti, of ^Mindoro^ is superficially one of the most Lastnva-likc. It 
has a short rhizomCj inarticulate stipes, a pinnate frond "with obscurely 
toothed pinna?, and glandular pubescence, but the glandular trichunu^s 
are very unlike those of La.slrmi, and the plant is distinctively Poly- 
podioid in the pale;t; and in the attachment of the pinmr. 
In my opinion, PoJypodium , in the broad s(*nse in wbich the genus ia 
usually construed, is more cumbrous than is justified by its unquestion- 
able naturalness. No group of whatever rank is justified except by 
naturalness (or by ignorance). The tribe is natural and among its 
various subdivisions can be seen every measure of consanguinity. A 
genus should be convenient as well as natural, and it ought not, within 
itself, to be more diverse than can well be avoided. Our Pohjpodia fall 
into four natural, readily recognizable, and definable groups: Pohjpodium 
in the narrower sense, Goniopldeljium, Phijiualodes, and Drynanop^Ls 
(P. he rack urn) ; to these Mynnccophila, but not SdVujaca, should per- 
haps be added. However, the scope of this paper does not include the 
settling of generic nomenclature where taste alone is at stake, and in- 
dividual initiative in such a matter may well await the joint action of 
pteridologists. I therefore follow Diels and Christcnsen,. so far as they 
agree. 
Within Enpohjpodium are several groups which in their full develop- 
ment are well characterized by the form of the frond, a character which, 
even if they did not intergrade, would but weakly justify their elevation 
to generic rank. Yet they are natural groups. The most probable 
central group is that with the fronds pinnate or nearly so, and superficial 
sori, representatives of which are P. minidum and P. macrum. From 
these, a series of forms can be found leading to Blume's Ctenopteris, 
P. celeUcum and P. ohliquatum, with very deeply immersed sori. From 
this latter group, as has already been shown, Prosaptia has been derived. 
Polypodium papillosum is a rather aberrant species, suggesting Gonio- 
phlehium in some respects, but not so much so that I would wish to 
combine them, nor would I treat it as a distinct genus {Thylacopteris 
Kze. ) . 
Polypodium irickomanoidcs is intermediate between P. macrum and 
the Calymmodon group, the best known representative of which is P. 
cucullaium. From this group, as has already also been shown, Arrosorus 
has been derived. 
Through such forms as P. solidum and iis variety denticulaia, or 
through species like P. loJierianum. and P. pleiosoroides, with irregular 
margfns, the P. macrum group is connected with the large section with 
entire fronds, called Grammltis by Blume, who recogni:^ed, however, that 
the sorus was not always elongate. When ferns in any group have 
