G9 
+ 
it is not^ as Lastraa is, primitive, and that it is not so particularly 
fitted to any set o£ widespread conditions — as rolypodiniu is to tlio 
epi]>liyiie habitat — that it can Leconie dominant under them and then 
become further specialized luider peculiar subordinafe conditions. 
Of our fern genera, only two, Scolopcndrium and StenocJdaena, can 
be treated at all positively as derived from Asplenium. The case of the 
former is already well known. If my S. pinnaium is what it is called, 
which I do not doubt, the genus Ttiphlchia owes \U origin to an error, 
for its sori originate exactly as in 8. vulgare. 
• AspJeninin epipliijticum. is a fern described from mntorial collected 
on the Gulf of Davao. It is common along the lower border of tlie 
I high forest at San Iiamou, and 1 have now specimens from Surlgao. lii^. 
resemblance to some other scandent AftpJrnin with pimple fronds was 
strong enough to demand some care in the diagnosis; yet ('hrist says 
it is unmistakably Sienochlaena, and,- indeed, its vegetiitive structures, 
root, stem, and leaf, are apparently identical with tltosc of oc(»nsioual 
immature forms of 8. aculeaia (Blume) Knnze.^" Knowing a fern as 
well ab I do this one, and never having seen any indication that it is 
other than Asplenium in its fruit, or has any other structures that as 
an Asplenium it might not have, I can only believe it to be an A.^pleniuw ; 
but neither can the affinity to 8tenocldaena be mistaken. In my opinum 
I we have in 8. aculeciia a very striking and perfect example of the repeti- 
tion in the development oftlie individual of the race-history of Stcno- 
I chlaena, proving beyond any doubt that it is an offshoot of Asplenium. 
The remarkably complete preservation of the race-history in this case 
is because the forms which are gone through are themselves well adapted 
to the environment. 
It is not impossible tliat Coniogmmme is also an offshoot of Asplenium^ 
but the evidence on which any particular ancestry might be ascribed to 
this genus and to Syngraiama is still inadequate. 
The Pteridece are so poorly represented in the Philippines that a 
discussion of the affinities of most of tlie group is not called for lierc. 
I have already pointed out the interesting position of 8rliizostegc as a 
probable ancestor of Fteris and Cheilanthcs, The genus was first de- 
■^ scribed by Hillebrand from a very rare Hawaiian plant. Baker ret^luced 
it to Cheilanthcs, and Clirist, with rather better reason, transferred it 
to rteris. My -Mindanao plants agree in every essential character with 
the Hawaiian, but in aspect incline toward Cheilanihes, ratlier than 
toward Fteiis, In the irregular and imperfect marginal anastomoses, 
Schizostege stands directly between Cheilanihes and Fteri.s. Its antiquity 
is attested by its discontinuous distribution. The resemblance of 8, 
Lydgatei to P. qnadriauriia suggests that the latter may be the most 
primitive forin of Pteris, and this idea receives some support from the 
"Underwood: The Ornus Strnochlapna. Bull. Torr. Hot. Cluh. (I!»OC), 33:40. 
