160 Rhodora [SEPTEMBER 
together 1.5-2 mm. long and 0.7-1 mm. wide, the perigynium about 
as long as the perianth or longer, usually bearing one bract on one side 
and two on the other; perianth conical, shorter than the bracts, 
irregularly plicate in the upper part, composed throughout of more or 
less elongated cells, the mouth minutely crenulate (or short-ciliolate) 
from projecting cells: androecium short, the bracts mostly in two to 
six pairs, imbricated, deeply saccate and complicate, the dorsal part 
usually a little smaller than the ventral but sometimes equalling or 
surpassing it in size, keel strongly arched throughout or becoming 
slightly concave in the outer part; antheridia borne singly or in pairs: 
mature sporophyte not seen. [Plate 126.] . 
On steep, damp rocks, usually in shaded ravines. Maine: banks 
of the Carrabassett River, Jerusalem (J. F. Collins 1609; listed by 
the writer as N. hyalina, RHopona 4: 209. 1902). New Hampshire: 
banks of the Ellis River, Jackson (A. W. E.; distributed in Under- 
wood & Cook's Hep. Amer. 83, as N. erenuliformis; listed by the 
writer as N. obovata, Proc. Washington Acad. Sci. 2: 298. 1900); 
the “ V," Waterville (A. Lorenz 203; listed by the writer as N. hyalina, 
Ruopoma 10: 192. 1908); same locality (A. W. E.); between the 
Greeley Ponds, Waterville (4. Lorenz); Profile Brook and the Flume, 
Franconia Mountains (C. C. Haynes, A. Lorenz & A. W. E.; listed 
by Miss Lorenz as N. hyalina, Bryologist 11: 113. 1908); Coosauk 
Falls, Triple Falls and the Ice Gulch, Randolph (A. W. E.); Hunting- 
ton Ravine, Mt. Washington (4. W. E.). Vermont: Downer’s Glen, 
Manchester (A. J. Grout; listed by the writer as N. obovata, RHODORA 
7: 58. 1905). Massachusetts: Mt. Greylock (4. L. Andrews). 
Connecticut: Beacon Falls (4. W. E.; listed by the writer as N. 
hyalina, Conn. State Geol. & Nat. Hist. Surv. Bull. 11: 51. 1908). 
New York: Rainbow Falls, Adirondack Mountains (W. G. Farlow). 
The specimen collected by the writer at the “V,” Waterville, New 
Hampshire, on August 18, 1911, may be designated the type. 
In some respects N. obscura is intermediate between N. hyalina 
and N. obovata and it is not surprising that it has been confused with 
both. It agrees with N. hyalina in its dioicous inflorescence and with 
N. obovata in its deep perigynium; all three species have in common 
the following features: suborbicular entire leaves, normally rounded 
at the apex; thin-walled leaf-cells with trigones, the marginal cells 
essentially like the others and therefore not differentiated to form a 
border; more or less abundant rhizoids, often pigmented with red 
or purple; and a plicate perianth without a beak, composed (at least 
in part) of elongated cells. 
