LINDMAN, AMERICAN SPECIES OF TRICHOMANES SM. 1< 



found it in the forests of southern Paraguay, tig. 1, B, Ex- 

 ped. 1. Regnell., n. A, 1803. (The South Brazilian Tr. Pab- 

 stiannm K. Müll., Bot. Zeit., 1854, p. 738, is closely allied 

 to Tr. hymenoides and perhaps only a form of it). 



2. Trichomanes sociale Fee. — Fig. 4—6. 



Fee, Cryptogames vasculaires du Brcsil, II, p. 85, tab. 85, fig. 3. 



I venture to maintain as a distinct species the plant, 

 which Fee has illustrated as stated above and called sociale, 

 anting specimens from Glaziou's collections n. 5242.^ I have 

 ^een numerous specimens of the same plant in Stockholm (Herb. 

 Regnell), and I find them very nearly allied to the prece- 

 ding, to which Prantl has referred this number of Glaziou's 

 <;ollections. The near relation to Tr. hymenoides is seen by 

 Fee's description: »les lames plus larges que longues», al- 

 though his illustration exhibits a more elliptical frond. 



I have examined this plant, 

 -I'arefully comparing it with Tr. 

 hymenoides (both the West In- 

 flian and Brazilian specimens); 

 the cellular structure is exactly 

 the same, fig. 4; involucre, cap- 

 sules, and spores are also almost 

 identical. The involucre of Tr. 

 ■sociale is perhaps somewhat smal- 

 ler, and is more distinctly wing- 

 ed along the sides (otherwise 

 free and exserted, as in Tr. hy- 

 menoides); and the lobes of the 

 moutli have a narrower edge, 

 consisting of 1 — 3 rows of flat, 

 rectangular cells with very thick 

 walls, fig. 6, C. (In Tr. hyme- 

 noides I always found a broad, easely broken edge or margin 

 of about () concentric rows or strata of such cells, fig. G, A, 



Fis- 5. Trichomanes sociale Fee. 

 upper half of a frond from Rio de 

 Janeiro, Glaziou n. 5242 (X 5). 



^ Fee writes »Glaziou 5249», but this is probably an erratum. 

 Arkiv för Botanik. Band 1. 2 



