580 Mr. Brown on Lyellia, Leptostomum, and Buabaumia. 
Better evidence on the same subject is afforded by Trichosto- 
mum, Didymodon, and Leucodon, in all of which the thirty-two 
teeth are distinct, though approximated in pairs; by the sixteen 
bifid teeth of Dicranum and. Fissidens; and by the like number of 
teeth with a perforated axis in Trematodon, Weissia nuda, Didy- 
modon latifolium, and several species of Grimmia. | 
In all the genera having a double peristomium I believe the 
pellucid axis more or less manifestly exists; but in these genera 
there is a great uniformity in the apparent number of teeth in the 
outer peristomium ; there being no instance of actual division in 
this series beyond sixteen, or of a further approximation, unless 
in Orthotrichum, in several of whose species the approximation 
or even union of the double teeth by pairs takes place, while in 
a few others the sixteen teeth are slightly divided at the apex, and 
in the whole genus the pellucid axis is remarkably distinct. 
The only exceptions to the actual division into thirty-two, or 
t.e structure indicating that number, in the simple peristomium 
of mosses, occur on the one hand in certain species of JFeissia; 
perhaps in Encalypta and in Octoblepharum, in all of which, I be- 
lieve, there is a reduction to sixteen: and on the other in Polytri- 
chum, where the number is frequently increased, varying in the 
different species, and chiefly by multiples of sixteen, from thirty- 
two to eighty. In this genus also, whatever the number may 
be, the teeth never have»a semipellucid, but rather an opake or 
thickened axis, and no tendency to union or even approximation 
is observable. The constant equidistance of the teeth in Poly- 
trichum seems to be connected with its peculiar mode of disse- 
mination; for as this takes place through the interstices of the 
teeth, and as complete separation of the seeds seems necessary on 
account of their extreme minuteness, a reduction in number 
and consequent increase of size of these apertures would. proba- 
bly, 
