Campbell's Islands] FLORA ANTARCTICA. 121 



to take place through the decay of the theca, when, as frequently happens, the theca and seta are together detached 

 from the plant, and possibly, if persistent, the operation may be aided by the development of a fungus which we 

 have seen in the walls of the capside of V. hyperborea. 



The more complicated structure of these parts, which most peristomed mosses possess, may eventually prove 

 mere modifications of, or deviations from the simpler organization of Voitia and Sphagnum. Gymnostomum pyriforme 

 tends to confirm such a theory ; in it the columella (what is considered as such being the inflected portion of the 

 sac), ascends from the bottom of the theca to the level of the stoma, and then, expanding outwards, like the 

 mouth of a funnel, reunites with the seminal sac around the rim of the stoma (vid. Grev. and Am. 1. c. vol. p. — ). 

 After a time, the edge of the funnel breaks away from that of the sac and with the columella shrivels up, thus giving 

 egress to the sporules. 



Mr. Valentine, in his ' Genera of Mosses,' has accurately described the cellular tissue of Sphagnum, which is, 

 in several respects, exceedingly curious. The cells themselves are bounded by very thick lines, formed of slender 

 tubes, running between the contiguous cells, but on one side of the leaf only ; a transverse section of a small por- 

 tion of a leaf, exhibiting both the cells and the interjected tubes, is shown at fig. 46. Valentine considers that the 

 latter are derived from the elongated tubes of the stem. 



The cells themselves are furnished, in some forms of the genus, with one or more spiral filaments, closely ad- 

 hering to their walls, sometimes these are entire throughout the length of the cell, at others broken or both broken 

 and branched. We are inclined to believe that the spiral filament is terete and adheres by a small portion only of 

 its surface to the tissue of the leaf; the extreme minuteness and transparency of the parts, however, increase the 

 difficulty of determining such a point with the accuracy that is desirable. No function has, hitherto, that we are 

 aware of, been assigned to these filaments ; they may act powerfully in enabling so delicate a tissue to withstand the 

 pressure of the water. 



The pores, by means of which a communication between the cavity of the cells and their surrounding medium 

 is preserved, are, in most cases, numerous and large, in others less so ; they appear more frequent on the upper sur- 

 face of the leaf, but are by no means confined to it, for sometimes they are placed opposite to one another, when the 

 leaf itself is perforated. They exist both in cells provided with spiral filaments and without ; in some instances, 

 where the. spires are broken and branched, the pores are bordered with a thick ring given off from the filament, 

 whence probably arises the supposition that what appeared to be pores were supplementary coils. They vaiy 

 greatly in size, occasionally extending completely across the cell. Valentine describes them as resembling a minute 

 truncated cone ; to us they appear on the same plane with the walls of the cells, except where their edges are thick- 

 ened, as described above. In S. niacropliyllum the cells are devoid of any filaments, are very narrow, much elongated, 

 and each perforated by from 8 to 14 large pores, which sometimes nearly divide the cell on one side ; they are 

 by far most numerous on the upper surface of the leaf. The uses of these pores are hitherto unknown, they may 

 be due to the nature of the situations in which the species are found. We have not proved them to reside in the 

 intercellular tubes, though their existence in their walls is possible also. On the other hand, the latter alone may 

 continue reservoirs for water during diy seasons, when, from the porous nature of the former, they readily part with 

 their moisture. 



Until the dehiscence of the operculum, no rupture of the calyptra takes place in Sphagnum. We have not ob- 

 served the desilient property of the operculum, noticed by some authors. The sporules in all the species are clustered 

 together in fours, of which three only are generally visible at first sight, as represented at Fig. VI. In most of the 

 species the surface of the theca is studded over with stomatiform pores ; these are however very inconspicuous, if 

 not wholly absent, in S. cymbifolium. The true pedicel is included within the vaginula and is singularly dilated be- 

 low the orifice of the latter. 



Plate LVII. Fig. VI. — 1, Longitudinal section of theca of S. cymbifoliwm ; 2, section of cellular substance 

 from centre of the same ; 3, spores in a very young state ; 4, cellular tissue of leaf showing spiral vessels and pores ; 



Y 



