26 BRITISH HEPATIC^. 



either entire or emarginate, vaginate, and embracing the base of 

 the involucre ; apex erect or reflexed. 



Involucre obovate (f. 8, 7), thick-walled, formed by the dilated 

 and hollow apex of the stem, and the combined bases of the two 

 involucral leaves, and amphigastrium. Frequently a lower leaf, or 

 pair of leaves, with their amphigastria, are attached to its walls, 

 proving that it is only a modified portion of the stem. The lobes 

 of the involucre are roundish-ovate (f. 8, 6) or cordate, sharply 

 emarginate, sometimes undulate, erect or connivent, imparting a 

 compressed appearance to the organ. The amphigastria, which 

 are connate with the lobes, are ovate, entire or sub-dentate, and 

 with reflexed margins. 



Colesule (f. 8, 7) immersed and adnate with the involucre, the 

 apex only free, cleft into five broad equal segments, but conoid (as 

 in N. hyalina, &c.) before the egress of the fruit. 



Calyptra obovate, with large pellucid reticulation, bearing on 

 its lower half 10 to 12 abortive pistillidia. 



Capsule oval (f. 8, 6, 7), chocolate-brown ; valves 4, elliptical, 

 of thick texture, beautifully annulate-striate. 



Feduncle ^" or more in length, rather stout and succulent, its 

 base sunk into the hollowed receptacle, and surrounded by four 

 fleshy lobes, a modified form of vaginula. 



Spores dark-brown, round or obscurely three-angled, epidermis 

 rough, -Yfto" i^ diameter (f, 8, 8). 



Maters long and flexuose, dark-brown •^" in length by j^^" 

 broad. 



Andrcscium terminal, spicate, on distinct shoots. 



Perigonial leaves exactly round, more concave, closely and 

 arcuately imbricated, bearing at their bases 2-4 conspicuous, dark- 

 olive, spherical or obovate anfheridia, seated on a slender stalk 

 about equal to their length. 



Sir W. Hooker states that tte leaves of this species, lite those of Mylia Taylori, 

 Gr. k B., "produce a minute, black, hispid, nearly spherical fungus," which is attached 

 to the surface or margin. 



In the barren state Nardia scalaris resembles closely some of the round-leaved 

 Jungermannise, but it may be recognized by the triangular-subulate amphigastria, 

 and the presence within the leaf-cells of the peculiar corpuscles already described. 

 When fructification is present, it is not likely to be mistaken for any other species, 

 except N. ohovata, in which the colesule is connate for half its length with the involucral 

 leaves, and N. hyalina, where the colesule is immersed. From both, however, it may 

 be known by the colour of the rootlets, which in N. scalaris are white, whereas in the 

 others they are purple. 



A revision of the forms named i^^ scalaris in my own herbarium (assisted by original 

 specimens fi-om Professors De Notaris and Lindberg), has convinced me that at least 

 three species are confounded under that name : — 1. The true N. scalaris, which is 

 dioicous ; 2. N. repanda, Lindb. ; and 3. N. geoscypha (De Not.), which are monoicous. 

 This discovery was made too late to allow of the species being engraved. Already the 



