60 
the costula of the secondary pinnula. The ultimate segments are generally broad 
oblong, obtuse, and the veins do no reach the margin. 
Naturally some of the bipinnatifid species dealt with in the first part of this 
monograph must be placed ind the present group, viz. D. honesta, yungensis, biserialis, 
leptosora, platyloba and D. hirsuto-setosa. They differ from the bipinnatifid species of 
the group of D. submarginalis by their furcate veins and pubescence with other hairs 
than Clenitis-hairs. The three first species together with D. macrotheca form a special 
small group, characterized by brown colour of the dried leaf; D. leptosora and 
D, platyloba are intimately related to D. vasta and it is a matter for conjecture whether 
these species are not less divided forms of D. vasta; it seems probable that several 
of the species of this group vary during life-time growing with age larger and 
becoming more divided. D. hirsuto-setosa is similarly a less divided species allied to 
D. Karsteniana. 
Concerning the sori we have within the group two series of species, one with 
exindusiate, the other with indusiate sori, and originally I made the two series 
distinct groups within the subgenus. A comparative study of all species will show, 
however, that the character indusiate or not can not be considered a principal 
one. Considering several other characters we find that the indusiate species are 
less related inter se than they are to exindusiate species. Thus D. villosa is certainly 
closely related to D. Karsteniana, D. andicola to D. atrogrisea, D. villosula to D. molli- 
coma and D. crenulans much nearer related to other Brazilian species than to 
D. villosa. Some authors, f. inst. JeNMAN, united into one species both the indusiate 
and exindusiate forms, but I have never found evidence that the very same species 
may be found with and without indusia; the presence or abscence of indusia is to 
me, at least as regards these species, a good specific character. It is, of course, 
rather difficult to use that character by the determination of specimens, as some- 
times all the indusia of dried specimens: have fallen off. 
In this group I place with some doubt the new species D. adenopteris, which 
comes very near to the ampla-group, as does also D. Iunensis; the xerophytic D. hirtula 
from Brazil looks very peculiar, still I think its best place is within this group. 
D. grandis differs from all other species by the cost above being perfectly glabrous 
without Clenitis-hairs. 
_ Working out the following key I have especially considered the scales and 
pubescence; Metrentus has (Fil. Lechl. II) given a key to the species known to 
him and there especially taken the Shape and structure of the scales into consi- 
_ deration; the characters pointed out by him are, however, not all constant, and I 
cannot use them. It is convenient to use the scales and pubescence in a key, because 
most of the specimens of the herbaria are very fragmentary, generally a single pinna 
of a leaf. Such characters as shape and degree of division, colour and structure of 
the basal scales are, therefore, of little value by the determination of such large species. 
The scales of the pinnz, their pubescence can, on the contrary, always be examined. 
