od filled with granular matter: so are the cells
throughout, to a greater or less extent. They are to be seen in all
stages of development on the pinnae of a very young frond, those near its
base having perhaps effected their purpose, while those at the apex of
the pinna, or the prolonging part of pinnula, may be formed of only one
cell. It is curious that the terminal cell does not become spherical for
some time: in its earlier stages it is cylindrical like the rest.

The appearances of the old ones are, if possible, more markedly in favour
of my hypothesis; there is the same aggregation of grumous _congealed_
matter about the ends of each cell, the same curious communication
between these masses which hide the septa from view, evincing a greater
or less tendency to assume the peculiar fuscesent or fusco-brown
appearance. I observed in two instances what appeared to me decided
irregular openings in the terminal cell, from one of which grumous
filaments projected; these appeared to communicate with the mass in the
terminal cell, which like that in all the others, is congealed; but it
assumes a different and very undefined form.

People may object and say, why were not more met with _opened_? This is
no objection, because it is obvious that a spherical body may be opened
in part of its surface, and yet unless this portion happens to be on the
_edge_ as it were of the sphere, it may escape detection with a
microscope of poor penetration.

In this the ramenta are confined, or nearly so, to the under surface of
the fronds. Most occupy that which is called the costa. In this the
first change as in Adiantum is in the definition of the margin. But this
point I have not paid much attention to, as with my present means here,
it would be absurd to attempt _proving_ how the fecundation takes place;
all that I can attempt is, to ascertain from structure and analogy, the
male nature of these curious bodies.

_See_ Plate _B_ for the various sketches. {450}

The next genus examined, is perhaps the instance in which these ramenta
have the strongest resemblance to ordinary simple hairs, both in their
young, when they represent succulent, tinged, grumous
molecular-containing hairs, and in the old, when they represent long,
flattened, coriaceous hairs, still there is abundant evidence to prove
that, however different these bodies are in appearance from those of
Cryptogamma, that they undergo the same changes, excepting perhaps