338 ON LYEJJLIA, LEPTQSTOMUM,
been able to detect ; but I observe in many cases an appearance which may perhaps account for the belief in their existence, namely a minute reticulation on the outer surface of the membrane, apparently owing to the cor- responding areolae of the inner surface of the operculum, with which it was originally in contact.
Polytrichim is remarkable for the various forms of its capsule. Those species, indeed, in which it is quadrangular have been considered as a distinct genus by Mons. de 568] Beauvois. But unless this difference of external form should be found connected with others, either in the habit or in the internal structure, which I believe is not the case, it seems hardly sufficient to justify the subdivision of so natural a genus.
The symmetrical quadrangular capsule, admitting of a regular cylindrical inner membrane, is a character of inferior importance to the plano-convex or dimidiate capsule, which almost necessarily implies, and in Daivsonia and Lyellia at least is found to be connected with, a corresponding irregularity in the figure of the cavity ; and hence I have introduced this remarkable form into the characters of both these genera.
On the structure of the inner membrane of the capsule in Poly iridium the only observation that I have met with occurs in English Botany, where, in the account of Poly- trichum subrotundum (plate 1624), it is stated that Mr. James D. Sowerbyhas discovered, in this species as well as mPolytricJium undulatum, a real membranaceous peristomium within the teeth, which, according to Sir James Smith, " gives quite a new idea of the generic character."
This inner peristomium, if it may be so termed, which is well represented in the figure referred to, and consists of a horizontal projection of the inner membrane immediately below its apex, will, I believe, be found to exist in all the species of the genus, and in some cases reduces the aperture of the inner capsule to half the size of the outer at the origin of the teeth. It is always, however, quite entire, and, according to my observations, is formed of a doubling of the
�� �
