446 MR. G. BREBNER ON THE 
their development, owing to their abundant granular protoplasm 
with the nuclei near the common point of contact of the cells ; 
ef. figs. 1, 2,3, at a. After the appearance of the schizogenous 
space their relatively small size would be alone sufficient to 
distinguish them, but they are still characterized by the abundance 
of their protoplasm and relatively large and conspicuous nuclei, 
although the latter no longer occupy the eccentric position of 
the earlier stages. Figs. 4,5, and 6 show early stages in the 
development of the split, which takes place at or near the centre 
of the group. It is quite likely that in fig. 4 the split took 
place at the point of contact of four cells (7. e. as seen in trans- 
verse section), and that subsequently radial divisions took place 
in the young epithelium. It is not at all uncommon, where the 
original meristematic group was composed of a large number of 
cells, to find two spaces separated by one or more of the cells 
(tig. 6, br.c.). As the schizogenous space widens, the surrounding 
cells continue to divide by walls radial to the centre of the 
canal, and thus constitute a continuous small-celled epithelium. 
Divisions in the epithelial cells other than radial seem to be very 
exceptional except when the canal is still quite young. 
Fig. 7 shows a fairly advanced young canal from a leaflet of 
Marattia cicutefolia, which is interesting as illustrating a later 
stage of such a condition as seen in fig. 6. This is not a double 
canal, as sections both above and below tke level of the section 
figured showed a single cavity. In the present case two cells in 
the same straight line have been drawn out through the increase 
in the size of the cavity. Owing to the shrinkage of the proto- 
plasm, the walls of these cells were very clearly visible. Such 
cells which locally extend across the cavity of a young canal may 
be called bridge-cells (br.c., figs. 6 and 7). As the cavity increases 
still further in size the bridge-cells, when present, get torn away 
from their attachment where it is weakest and project further into 
the cavity than their sister cells. In one case, in a young canal in 
the root of Marattia attenuata, a remarkable example of accommo- 
dation to tension was observed. The original wall of a bridge- 
cell had been ruptured transversely to its long axis, and within 
this was seen the much-elongated protoplast of the bridge-cell, 
of uniform thickness throughout its length and clothed with a 
new wall. These bridge-ceils, which can only exist as such when 
the canal is still young, show no signs of mucilaginous dege- 
neration, their walls and contents being evidently in the living 
