Page:Miscellaneousbot02brow.djvu/357

 AND BUXBAUM1A. 341

necessity for this complete dispersion in Mosses seems to be inversely as the size of the seeds. For in those genera of the order in which the capsule either bursts irregularly or has a naked mouth, the seeds are in general larger than in those with a single, and still more manifestly than in those with a double, peristomium. And in conformity with this also, in Polytrichum undulatum and Icevigatum the tympanum is sooner ruptured or removed than in the other species of the genus.

The result of this comparison of Polytrichum with Ly cilia and Dawsonia, although it confirms the propriety of their approximation, does not afford any clearly distinguishing mark for the very natural section of the order which these three genera form. In the mean time, however, it may be circumscribed, though not with absolute precision, by a com- bination of the more general characters which have been now enumerated.

LEPTOSTOMUM.

In denning this genus, which was first proposed in my former paper on Mosses 1, I relied chiefly on the undivided annular projection of the inner membrane of the capsule. I was induced to employ this modification of the peristomium as a character, though certainly far from being obvious, in finding it to exist in several mosses of the southern hemi- sphere, having a similar and peculiar habit : and which, had it been neglected, I must have referred to Gymnostomum, with the greater number of whose species they have hardly anything in common.

Mr. Hooker, however, has since found, as he states, [67a the same structure in several other mosses, generally con- sidered as having a naked peristomium, particularly in Gymnostomum microstomum, GviffitJiianum, and fasciculare ; and as these species have but little resemblance to each other, and still less affinity to Leptostomum, he has reduced this genus also to Gymnostomum.

If the statement referred to be correct, Leptostomum,

1 [Vol. i,p.'d52.)

�� �