120 FLORA ANTARCTICA. [Auckland and 



They are roundish, on pedicels as long or longer than themselves, generally solitary in the axils of the perigonial 

 leaves, which are somewhat ventricose at the base, but not otherwise different from those of the stems. 



The anatomy of the theca of Sphagnum is very peculiar. What appears the columella does not extend, as in other 

 genera, to the summit of the theca, but is a continuation of the seminal sac, ascending from the bottom of the theca, 

 and forms a portion of the same membrane which also lines the under surface of the operculum, passing completely 

 across the stoma, as shewn in Plate INTL. fig. VI., where the central figure is drawn from a sketch and section pre- 

 pared by Mr. Wilson. 



This curious structure of the seminal sac is quite different from what obtains in most genera of mosses, and 

 appears to have been misunderstood by Arnott and Greville, in whose excellent essays upon the ' Genera of Mosses,' 

 the columella is described as sinking, along with its opercular membrane, so low, as to assume the appearance of a 

 tympanum, stretched across the interior of the theca, a little below its base (Weru. Trans, vol. iv. p. 181.); then figure 

 however does not represent the columella bearing any residua of the opercular membrane, but merely having the base 

 of the seminal sac drawn up into the axis of the theca in the form of a cone, which is its true origin. The more 

 striking peculiarities of the sporular sac of Sphagnum are these ; 1st. its forming a bag or cyst without any orifice : 

 2nd, in the drawing up of the base of this bag into the axis of the theca, but not so far as to reach the level of the 

 stoma, nor consequently the upper surface, or that opposite the base, which remains entire and stretched across the 

 stoma. If the columella were earned up to the same height as in other mosses, an obliteration of the upper part of 

 the sporular membrane woidd be caused by the perforation of the latter, (if we regard the metula as a portion of the 

 columella), or else there would ensue a mutual cohesion of the membranes of columella and sac. 



Sphagnum may be considered to possess the simplest form of sporular sac, the dehiscence of which is probably 

 caused by a removal of the upper portion in the same plane as the stoma and parallel to the operculum. The next 

 stage of development of this organ is, perhaps, presented in some astomatous mosses ; Voitia *, for instance, a per- 

 pendicular section of the theca in which genus exhibits the seminal sac in the form of a vertically elongated ring, 

 supported in the axis of the theca by the corculum of the columella. The latter passes uninterruptedly from the 

 apex of the seta to the top of the persistent operculum, thus apparently perforating the sac, by whose inflected walls 

 it is lined for the greater part of its length. In this case, dehiscence and the escape of the sporules may be supposed 



* In the young state of Voitia hyperborea, of which (through the kindness of Sir James Ross), I have exa- 

 mined many thecas, there is a communication between the seminal sac and the lining of the walls of the theca (thecal 

 membrane), by means of conferva-like filaments such as are seen in most other mosses. Tracing the different 

 membranes upwards, from the apex of the theca, I was led to believe that the same tissue formed the thecal mem- 

 brane, the conferva-like filaments, and the corculum of the columella ; and further, the immediate communication 

 between all the surfaces of the seminal sac and the walls of the theca afforded room for a conjecture, that the latter 

 were immediately concerned in the development of the sporules, especially as before the separation of the spores 

 both the thecal membranes and filaments were full of a grunious fluid, which afterwards disappears. If such a view 

 be correct, the internal structure of the theca of Voitia is very simple, and consists, 1st, of stout cells forming the 

 external walls ; 2nd, of a fine tissue, not only lining the former and sending filaments to the opposite walls of the 

 seminal sac, but, becoming more condensed at the base and apex of the cavity of the theca, it ascends in its axis and 

 meets the descending portion in the hollow of the columella, over whose surfaces they both ramify ; and 3rd, of a 

 vertically elongated ring (the seminal sac) through whose centre this second membrane passes. Mr. Wilson (to 

 whom I am indebted for my knowledge of the structure of both theca and cellular tissue of Sphagnum), has not been 

 able, from the want of specimens, to confirm this view of the structure of Voitia ; he, however, informs me, that 

 " the existence of such a connecting tissue is only analogous to what is found in Gymuosttomum pyriforme, whose 

 sporular sac is connected with the base of the theca by a bundle of filaments;" he also adds, that " in many Poly- 

 tricha there is a distinct, almost woody, central axis to the columella, with filaments intervening betwixt it and the 

 winged folds of the sporular sac which forms the outer part of the columella." — J. D. II. 



