aes 
Subgenus IX. Ctenitis ¢. Chr. 
To this distinct subgenus (or genus) belong most of the decompound species 
of tropical America. I have little to add to my description of it in the first part, 
though I must add the most important character: the spores are always echinate, 
Several of the decompound species have only short, red-brown articulated hairs, 
which can always be found on the cost above (see, however, D. grandis); by these 
two characters: hairs and spores, the subgenus is absolutely different from the two 
former, Eudryopteris and Stigmatopteris. Also the scales are characteristic, generally 
clathrate. In the group of D. subincisa other hairs than the common Cienitis-hairs 
occur; the costules of the species of that group are above strigose by antrorse, 
subulate hairs, which are paucicellular and only a modification of the short Cfenitis- 
hair, although not always distinctly articulated. Frequently the ordinary Ctenitis-hairs 
are to be found under the subulate ones; besides these hairs also others may be 
found, long, flexible, pluricellular or short, unicellular. From the subgenus Para- 
_ polystichum (D. effusa and allies), where articulated hairs also are found, the species 
of Ctenitis is different by the structure of the lamina, which is cyatheoid (see 
introduction). 
All species of the groups 1—4 have an erect or oblique rhizome with fascicu- 
lated leaves, the fifth group, which includes a single American species » only, 
D. protensa, differs from all others by its creeping rhizome with scattered leaves, 
that are are more or less tripartite; that species is thus very different from other 
American species, still it shows all important characters of Cfenitis. 
I now know 66 American species of Ctenitis, of which 25 were dealt with in 
the first part. As shortly pointed out there (pag. 84—85). the species may be sorted 
into some rather distinct groups, and I now think the five groups characterized in 
the key below may be considered rather natural and well defined, although there 
are found some species that with equal right could be placed in two groups; such 
species are considered in the keys to the species of both groups. : 
Key to the Groups. 
lanceolate or deltoid, not tripartite (see, however, D. Hemsleyana). : : 
— Rhizome creeping with scattered leaves. Lamina generally distinctly tripartite ; 
hairs all common Cfenitis-hairs. 5. Group of D. protensa, “po 
1. Rhizome erect with fasciculated leaves. Lamina pipinsiatioe: qaednpesneute. 
