pA 
As far as I can see it is hardly possible to distinguish these two forms from 
each other; the extremes described are connected by all possible intermediate forms. 
D. grandis is apparently confined to the mountains round Rio de J aneiro 
(Corcovado, Tiyuca, Serra dos Orgios) and it is gathered by nearly all collectors, f. inst. 
GLaziou nr. 966 (H =f. Fischeriana), 1676 (H, Rg), 5299 (B), 5385 (B, H, Rg), 5386 (B, H); REGNELL 
nr. 256 (Rg); Mosén nr. 2695 (B, H, L, Rg, S, U) and others. 
336. Dryopteris connexa (Kaulf.) C. Chr. Ind. 258. 1905. 
‘ Syn. Polypodium connexum Kaulf. Enum. 120. 1824; Mart. Ic. Cr. Bras. 90 t. 65. 
: papier connexa Feée, oe 243. 1852, Cr. vase. ee 1: 101; Mett. Pheg. nr. 
61; Fil. Lechl. 1, 
Palepolianh acriniecan Y connerum Bak. Fl. Bras. 17: 502. 
Phegopteris scrobiculata Fée, Cr. vasc. Br. 1: 102 t. 31 f. 2. 1869. 
Phegopteris eriopodia Fée, |. c. 102 t, 31 f. 1. 
Phegopteris adnata Fée, |. c. 103. 
? Phegopteris propinqua Fée, |. c. 108 t. 32 f. 3. 
Polypodium Willsii Bak. Ann. of Bot. 5: 458. 1891! 
Dryopteris Willsii C. Chr. Ind. 301. 1905. 
Type from Brazil: Santa Catharina leg. CHamisso (not seen, but a specimen 
in B is said to agree exactly with the fragmentary original specimen in Herb. Kaulf. 
no. 879 (on a label by Merrenivs). I, therefore, consider that specimen typical). 
ce 2 Sua eo Fe, Fo ee 
a 2 
An extremely variable species, if the numerous forms referred to it really 
belong here; scarcely two specimens are exactly alike. Still I cannot find good 
characters by which they may be distinguished. It varies especially in size, degree 
of cutting and cutting. The typical specimen mentioned above may be described thus. 
Lamina tripinnatifid; pinne at distances of 8 cm, stalked, acuniinate, about 
25 cm long. Basal pinnules sessile, the middle ones decurrent and connected by a 
wing, 6—7 cm long, 1 cm broad, shortly acuminate; lower basiscopic pinnules of 
basal pinnze somewhat longer, otherwise not different. Pinnules lobed about halfway 
down to the midrib into truncate, somewhat arcuate lobes. Texture chartaceous, 
colour when dried brownish. Rachis and costze beneath very sparsely fibrillose with 
| linear, hairlike, dark-brown, toothed scales. Coste and costules above antrorsely 
x strigose with subulate, rather long hairs; the veins above with similar, scattered sete ; 
ee margins ciliate; costa and costules beneath laxly pubescent with articulated, rather 
| soft, long, whitish hairs; short hairs none. Veins subpinnate in the lobes, not 
reaching the margin. Sori 4—6 to a lobe, about medial, exindusiate. 
The rhizome, as shown in more complete specimens, is erect, thick, densely 
> coated with soft, linear, ferrugineous, sharply toothed scales, which are more or 
less wool-like and intertangled; Merrentus (Fil. Lechl. II. 26) described them as 
being margined by thicker cells, which I, however, cannot see. 
D. connexa differs from D. grandis by its tripinnatifid lamina and by its s strigose 
et Fanner Oo te ae ea he EES Sala Pe 
wae Aire ice ecinee e 
