132 
The work was undertaken at the suggestion of Mrs. Elizabeth 
G. Britton, to whose kind criticism and assistance is due much of 
whatever value this revision may possess. These studies were 
begun about the time that Limpricht’s treatise on the Isotheciaceae 
(Rabenhorst Kryptogamen Flora) was issued and it was planned 
to: parallel his work on the European plants of this and related 
groups by a somewhat similar treatment of the American forms. 
Except the recent issue of the first part of Braithwaite's work 
on the pleurocarpous mosses, all the extended treatments of the 
Musci UN have followed Schimper's classification in prin- 
ciple. 
A somewhat fragmentary publication like this has followed 
the more extended treatises in this particular by reason of the com- 
paratively narrow limits of the work. 
This study was begun with no bias either for or against 
Schimper's classification, but, as the work progressed, it became 
more and more evident that the Isotheciaceae do not form a natural 
group, but a highly artificial one, being based principally on the 
following characters: Erect capsule, inner peristome attached to a 
narrow basal membrane, and absence of cilia. 
It is very easy to see that the cilia, being merely thickenings of 
cell walls, would be much. more likely to disappear wholly or 
partially as a variation than that the variation should modify the 
. general structure of the plant and leaf. In some plants the num- 
ber of cilia varies from one perfect cilium and another imperfect to 
three perfect cilia. In BracAythecium Fendleri, which is described 
as having cilia solitary and short or none, other capsules from 
type specimens possess two well developed cilia. In species where 
cilia have never been known to develop, the lack of cilia is of 
greater import, but even when associated with an erect capsule 
cannot be considered as a character suitable to distinguish families, 
although these characters may well distinguish species or even 
Pn their importance depending solely on their constancy. 
- Any one who will take the trouble to compare carefully the 
European species of Homalothectum with several species of Campto- 
?hecium- cannot help seeing that the two genera are much more 
closely related than Homalothecium and Entodon, or any genera of 
the Isotheciaceae. 
