120 
NEW ARRANGEMENT OF 
nion : " Toute pyxidule {theca nobis) est d'abord com- 
pletement solide et charnue; mais lorsqu'elle a pris un 
certain accroissement, sa substance interne se grenelle fine- 
ment, se distingue peu a peu du test) et passe enfin a Fetat 
d'*une tres fine poussiere. Assez souvent une portion cen- 
trale de cette substance persiste sous la forme d'un axe 
longitudinal, qu'on appelle columelle 
AVe do not by any means intend to assert that the spo- 
rulae are not in some cases formed in much greater abund- 
ance than in others : this we know to be true. The cellular 
matter, in which the sporulse are produced, if any remains, 
will also shrink, and perhaps adhere to the columella, par- 
ticularly if the sporules are first formed in that portion of 
the granular mass nearest the lining of the theca. In 
many cases, the columella has assuredly little appearance 
of a regularly formed body, and in some it is scarcely pos- 
sible to discover it at all ; yet when M e see the same part 
constantly exhibiting a beautiful and symmetrical structure 
in other species, it is surely fair to draw the general infer- 
ence, that it is not formed by the contraction of the mere 
debris of the cellulose or pulpy nidus of the sporules ; wit- 
ness the columella of Gymnostomum XantJiocarpurriy of 
G. pyrtforme, and G. involutum ; the columella of Daw- 
sonia, of Lyellia, of all the Polytricha ; of Tayloria^ Sys- 
tylium, and all the Splachna. 
The columella is also, in very many instances, evidently 
tubular, which could not be the result of contraction, but 
rather a proof of the contrary, at least in the sense of the 
above named learned Bryologists. We may here, however, 
repeat what we have stated, when mentioning the apophy- 
sis, to be our opinion, that, at an early period, the cavity of 
* Bulliard's Diet. Element, de Botanique, edited by Richard, p. 67. 
