318 OBSERVATIONS ON THE 



friend, l\h. Trancis Bauer, as well as from some observations 

 of my own, I am disposed to believe that considerable 

 diversities may exist in the placentation of mosses : that in 

 some cases the seeds may be formed in a much greater por- 

 tion of the colamnula than in others : and it is even not 

 improbable that in certain cases its whole substance may be 

 converted into seeds ; or, to speak more accurately, that it 

 may produce seeds even to the centre, and that the cells in 

 which they were probably formed may be re-absorbed. This 

 I am inclined to think is the case in Fhascum altermfoliiim 

 of Dickson, in the ripe capsule of which there is hardly 

 the vestige of a columnula ; and I have observed the 

 same structure in two new species of Anodontiinn of Bridel ; 

 which, if it equally exists in the only species of this genus 

 hitherto described, would perhaps considerably strengthen 

 its character. In these cases the inner membrane is also 

 316] evanescent ; and such a structure, it may be remarked, 

 equally militates against M. Beauvois' theory, whether we 

 suppose the columella to have existed at an earlier stage, 

 in the usual form, or not. 



As to this organ being tubular, and discharging its con- 

 tents by the top, it is neither consistent with what has been 

 already observed, nor with the appearance of its remains in 

 the ripe capsule : but, admitting for a moment its tubular 

 nature, there are certain mosses in which no discharge could 

 possibly take place in the way described ; the column being 

 elongated even to the apex of the operculum, to which it 

 often continues to adhere, as in Buxhaumia^ and in the first 

 of the two new genera which I now proceed to describe. 



DAWSONIA. 



Peristommm penicillatum, ciliis numerosissimis capillari- 

 bus rectis sequalibus e capsulae parietibus columellaque (!) 

 ortis. 



Capsula hinc plana, indc convexa. 



CalyiJtra exterior e villis implexis, interior apice scabra. 



Muscus liinc arete ciffinis Polytricho, quocum foliis^Jloribus 



