48 COPELAND. 
my work on the Polypodiaceae of San Ramon,* in which I carried 
the* natural arrangement of the family farther than had before 
been attempted, I was confronted by the unnaturalness of the 
supposed near relationship of this fern to Polypodium, and, 
not knowing what its real affinities might be, did not include it 
at all in the "family tree" of the Polypodiaceae. 
It has since become clear that Taenitis belongs in the only 
large and nearly natural tribe not represented among its com- 
panions in the standard works, the Davallieae. In the family 
tree just cited, Microlepia represents the central group in this 
tribe. By removing Davallodes, I have since made it possible 
to constitute a really natural genus, including Microlepia in 
Dennstaedtia. Deniistaedtia and its nearer relatives are un- 
failingly characterized by hard, creeping, reddish rhizomes 
clothed with small, harsh, reddish hairs which are pluricellular 
at the base but narrowed above to a single row of cells. The 
<ioIor and form of a negro's hair do not testify to blood relation- 
ship more certainly than do the color and form of the pubescence 
of Dennstaedtia, Sacccdoma, Leptolepia, Odontosoria, Tape- 
inidium, Saccoloma, and Taenitis. Trichomes enough like these 
to betray relationships are found not only in various Davallieae, 
but also, as we shall presently see, in various ferns which have 
apparently not been suspected of such affinity. 
Of ferns known to me, the most like Taenitis blechnoides 
(Willd.) Sw. in superficial aspect is Schizoloma ensifolium 
(Sw.) J. Sm. This fern is extremely variable in its venation, the 
veins being sometimes almost free, sometimes anastomosing 
very similarly to those of Taenitis, whose venation is likewise 
far from uniform. Taenitis sometimes, though not usually, has 
the peculiar form of leaflet, broadest near the base, characteristic 
of Schizoloma ensifolium. The rhizomes and the bases of the 
stipes are so alike in the two ferns that these parts are often 
practically indistinguishable. A young plant of either is likely 
to be mistaken for the other. The difference in the position 
and protection of the sorus is hardly greater than that found in 
Vittaria, between V. scolopendrina and V. elongata, where it is 
not usually regarded as constituting a generic distinction. 
Taemtis itself is variable in the position of the sorus, as the 
accompanying photograph shows (Plate II). And there are 
species of Schizoloma whose sori are by no means marginal. 
However, internal structure affords in general a better clue 
to affinity than does external ; and the steles in the rhizome of 
*This Journal 2 (1907) Bot. 74. 
I 
