618 FILICALES. 
a careful revision, with special reference to their anatomical characters, 
before this progression can be regarded as established. 
MATONIA-DIPTERIS SERIES, 
The genus Difreris, Reinw., so long merged in the comprehensive genus 
Polypodium on account of its sorus being naked and superficial, has recently 
been restored to its independent position, and is now held to be the sole 
representative of the family of the Dipteridinae.1 There is little doubt 
that this position is justified, while among relatively primitive types the 
family finds near allies among the Matonineae. 
The genus is represented by four living species from the Indo-Malayan 
Flora, which illustrate an interesting progression in leaf-architecture. They 
all have creeping rhizomes, showing occasional dichotomy, the type of shoot 
being closely similar to that of MJatonia. The axis and the bases of the 
leaves alike are invested with a dense covering of hairs, which are, however, 
flattened into elongated scales, an advance upon the filamentous hairs of 
Matonia. ‘The leaves of the different species vary in area, but are alike 
in plan: upon the end of a long petiole is borne a lamina which is repeatedly 
branched in a dichotomous manner. The branches may remain narrow, with 
a marked midrib and lateral flanges of no great area, as in D. Lobbiana, 
Hooker, and D. quinguefurcata, Baker: or they may be broader, and be 
more or less webbed into a lamina, which is, however, still divided by a 
median sinus into two symmetrical halves: this is seen in D. conjugata, 
Reinward (Fig. 343 A), and D. Wallichit, Hook. and Grev. This leaf- 
structure is comparable with that of AZazonda, in which also the outline of the 
lamina is referable to a dichotomous branching, and as in that genus, so 
here also the sori are in the narrow-lobed species disposed upon the flanged 
wings on either side of ¢he midrib. Their relation to the area of the leaf- 
surfaces within this very natural genus is instructive for comparison with 
other Ferns. In the narrow-leaved JD. Loddiana, and especially near to 
the bases of the several lobes, the sori form a regular linear series on 
either side of the midrib (Figs. 344 and 343 E) In D. guinguefurcata the 
lamina is larger and the segments broader than in D. Lodéiana, and the 
-areolae within the veins on either side of the midrib are larger, and contain 
more sori: these illustrate various degrees of fission, and thus they become 
spread over the enlarging area (Fig. 345). D. Wadllichit appears to occupy 
a middle position between these species and the large D. conjugata; for 
it is described as having the ultimate segments linear in form, and the 
sori as being similar to those of D. conjugata, but more numerous than 
in D. Lobbiana or guinguefurcata. Lastly, in the large-leaved D. conjugata 
the bifurcate lamina is broadly webbed, and the very numerous small 
sori, which are distributed over the wide expanse, may be circular or oval, 
.and not always distinct from one another: they may vary much both in 
1Seward and Dale, Phil. Trans., vol. cxciv., p. 487. 
