92 
smaller, deciduous. The ‘scales of the rhizome are brighter than those of D. protensa. 
The differences here pointed out are as a rule found between the variety and D. 
protensa but they are sometimes scarcely observable. Kiorzscn (loc. cit.) mentions 
a difference in pubescence: D. protensa has articulated hairs, the variety not; this 
difference does not exist; both have all ribs rusty-pubescent with common Cfenitis- 
hairs, and hairs of any other kind are never found. D. protensa has, it is true, 
generally several longer articulated hairs on the ribs above, which are not found 
in most specimens of yar. funesta, but they exist in younger, well-preserved fronds. 
. D. protensa with its variety is so well-known and so often described and 
figured that a new description is unnecessary. By its creeping rhizome, quin- 
quangular lamina and broad, obtusely rounded segments it is abundantly different 
from all other American species of Dryopteris. It is clearly a member of the sub- 
genus Ctenitis, within which with its African allies it forms a separate section. I 
hope I shall later on be able to dealing in some detail with this section and to 
prove its real affinity. 
D. protensa var. funesta is just as variable in America as is the true D. protensa in 
Africa. It varies in size and especially in degree of division. Some specimens have 
the middle pinnz pinnatifid only with all secondary pinnule decurrent and con- 
fluent at base and often fully entire or at best slightly crenate or lobed at the 
anterior side. From this, the most simple form, the basal pinne of which are 
nearly as long as the central part of the frond, all possible transitions are found 
to large forms with longer central part and with the middle pinnz bipinnatifid with 
all pinnule free and nearly or quite incised to the midrib into broad tertiary seg- 
ments which again are sometimes crenated or lobed. Such large forms are found 
in about all localities together with the smaller ones. A still more divided form is 
mentioned below. Aspidium acrocarpon Fée differs a little from the common form 
by its sori being nearly terminal on a vein, which, however, occurs occasionally 
in several specimens. 
D. protensa var. funesta has its head-centre in the Guiana-lowlands and Trinidad, 
where it is a very common fern; it goes south to North-Brazil and is besides found 
in much scattered localities as Guadeloupe, and Martinique, Sto. Domingo and Panama. 
Sto. Domingo, Mayernorr (B = A. acrocarpon Fée, leg. Tussac, not seen). 
Guadeloupe, L’HEeRMinier nr. 160 (B, RB), Mazé nr. 305 (B), Pére Duss nr. 4053 (RB, W). 
Martinique, Pére Duss nr. 1583 (RB). 
Trinidad, Fenpter nr. 16 (B, Rg, RB, W), Bot. Gard. Herb. nr. 140, 322, 323, 3757, 6058, 6282 (B, R, W), 
Broapway, nr. 2850, 2851, 2313 (RB), Hircucock nr. 10385 (W) and others. 
Guiana, British G., ScHompurcK nr, 16 (RB), 1178, 1203 (B), Appun nr. 10 (B), JeNMAN nr. 402 (B, RB). 
— Dutch G., Hosrmann & Kappier ed. HOHENACKER nr. 64°(B, RB), J.KnyPer (W). — French G., 
‘Leprieur (B, W), Sacor nr. 725, 725 bis (B, RB), Ricw (CE). 
Panama: Canal Zone, Maxon nr. 4896 (CC, W). Cana, Wittiams nr. 866 (W). 
