��AND BUXBAUMIA. 335
different states, to ascertain absolutely the oeconomy of this moss in the dispersion of its seeds. It is evident, however, if the description already given be correct, that as far as dispersion takes place, by the month of the capsule, it can only be effected by a contraction or shortening of the colu- mella, and a corresponding retraction within the cavity of the capsule, of the central portion of the peristomium which adheres to it ; and in this state several of the specimens were actually found.
But it is also manifest, both from the great size of the columella, and its numerous points of connection with the inner membrane, that in this way the whole of the seeds cannot be discharged. It is possible therefore, that in cer- tain circumstances at least, their dispersion may be assisted by the minute pores existing at the base of the capsule. These pores, in the specimens examined, were found to penetrate the outer membrane, or even its external layer only, and being situated below the origin of the seminal bag, their first effect will probably be to accelerate the decay of the internal spongy texture of the base of the [565 capsule. But by thus removing support from the colu- mella and inner membrane, they may contribute to the greater shortening of the former, and consequently in increasing the dispersion by the mouth of the capsule ; or, from the same cause, the inner membrane being at length ruptured, the seeds may be in part discharged by the pores themselves.
It seems then not improbable, from what we at present know of the structure of Lyellia, that for the dispersion of the seeds in this genus there are two distinct contrivances, both of which, however, in the only species yet known, are apparently imperfect. But hence it is perhaps allowable to conjecture, that either other species of Lyellia, or a nearly- related genus may exist, in which while the mouth of the capsule remains absolutely shut, the pores of the base may be sufficiently enlarged for the complete performance of this important function.
Pores exactly resembling those of Lyellia crispa have not hitherto been found in any other moss. I have observed,
�� �
