23 
Kaulfussia are occasional neighbors of the latter. Scattered through 
the forest at lower altitudes are Aspidyiim difforme, A. decurrens, Di- 
plazium bulbiferum and -Pteris quadriaurita, beside the strongly dimor- 
phous species, Psomiocarpa apiifolia, Stenosemia aurita, S. Pinnata, and 
Leptocliilus latifolius, and the subdimorphous Nephrodium diversilobum ; 
and Lindsaya gracilis, Syngrarnma alismcefolia, Diplazium pallidum, T). 
tomentosum (?) and Pteris pluricaudata, near the rain forest. The last 
is intermediate, sometimes growing with Diplazium Williamsi. 
The dimorphism mentioned above is, as far as it goes, a tropopliytic 
adaptation. Asplenium subnormals and the little Ophioglossum growing 
on creek edges are more completely tropophytic, usually disappearing 
during the dry season. Unless Adiantum philippense is one, I found 
no specialized terrestrial tropophytes in the savanna-wood at San Ramon. 
Elsewhere in the Islands, N othochlccna densa, Cheilanthes tenuifolia and 
Helminthostachys are parang plants active during the wet season only. 
The best defined society in this group of plants is that of the narrow 
ridges in the upper part of the high forest. These have a very limited 
flora, in which Gallipteris cordifolia and especially Taenitis and Poly- 
stichum aristatum are characteristic. Syngrarnma and Lindsaya gracilis 
are the only other ferns likely to be found here. 
The characteristic feature modifying all epiphytic vegetation is the 
limited and uncertain water supply. The structure and life of the plant 
are so profoundly altered in adaptation to this general feature, that 
details: of the environment, especially those in regard to the moisture 
present in the latter, are able to exert much less influence on either the 
structure or the distribution of epiphytes than on those of terrestrial 
ferns. Therefore, although many of the epiphytic ferns are confined 
to the % canons and only a few have not been found in them, a division 
of the epiphytes along the lines adopted in treating the terrestrial species 
would encounter too many doubtful cases and would be supported by 
altogether too little difference in structure to be justified. For the same 
reason, the epiphytic vegetation of the high forest is more closely related 
to that higher up the mountain than is the terrestrial. 
Epiphytic plants are those growing essentially without contact with 
the soil. They are so called because they usually grow upon other plants, 
but some flourish indifferently either in such a situation or on rocks, and 
a few grow as a rule, or it may be entirely, on cliffs or bowlders. Davallia 
pallida I found but once and then on stone. Such species might tech- 
nically be called lithophytes or petropbytes, but such a distinction would 
express no important biological difference. Species often or usually on 
rocks are Nephrolepis laurifolia, Asplenium vulcanicum, Antrophyum 
semicostatum (sometimes, even terrestrial), Ilymenolepis spicata, Nipho- 
bolus nummularicefolius, Polypodium revolutum, P. nigrescens, P. com- 
mutatum, P. albido-squamatum, Photinopteris speciosa, and the Davallia 
already cited. As rock-dwellers, these are practically confined to the 
