118 ON THE ANATOMY OF 
is at its maximum in Z. acanthocephalus, in which the tubes 
are smaller than the spokes. 
Sir W. Hooker has given a representation of the trans- 
verse section of a scape of the E. decangulare. Bot. Mag. New 
Series, t. 3126, in which I imagine the green reniform spots, 
are intended to represent the tubes. The central circle there 
represented I do not quite understand, since ће connection of 
the spokes, (which are a continuation of the central system), 
with the cuticle, is in all the species I have examined, unin- 
- terrupted and most distinct. 
With regard to the transverse septa, I am unaware that 
any opinion as to their nature and origin has been given. 
M. Mirbel indeed in a note on Marchantia, appears to 
think that they are орепіпоз іп the sides of the cells, and in 
particular refers to the septa of the plumule of Nelumbo, in 
which he says the clefts are so multiplied, that the cells are 
actually transformed into reticulate tissue. Examined in their 
perfect state, they do certainly appear to be clefts. In reality 
they are nothing but intercellular spaces, modified in form 
by the transverse distention they have undergone. That 
such is their origin is evident on examining the latest deve- 
loped parts of the leaves of Eriocaulez, or the peduncles of 
those species which have incomplete septa, we here find that 
the partitions are not flattened, that they are composed of vesi- 
cles of exceedingly irregular shape, containing green-globules; 
that these cells present two, three, or several gibbosities, by 
means of which they are mutually united, while they are to- 
tally unconnected elsewhere. Spaces will hence necessarily 
exist. We have then to assume such a degree.of lateral dis- 
tention consequent on the increase of diameter of each tube, 
that the cells become flattened, and the meats generally as- 
sume an angular form. That this distention operates in this 
way there can be no doubt, since longitudinal distention is 
obstructed by the septa being attached transversely to the pa- 
rietes, it can hence have only the effect of increasing the din 
