borne upon erect, elongated filaments, originating
from the creeping mycelium. These threads are hollow, and rarely
septate; the upper portion divided into numerous branches, and these
again are subdivided, the ultimate ramuli each terminated by a single
conidium. This body when mature is oval or elliptical, filled with
protoplasm, but there is a diversity in their mode of germination. In
the greater part, of which _P. effusa_ may be taken as an example, the
conidia have the function of simple spores. Placed in favourable
conditions, each of them puts forth a germ-tube, the formation of
which does not differ in any essential point from what is known of the
spores of the greater part of fungi.

The short oval conidia of _P. gangliformis_ have little obtuse papillae
at their apex, and it is at this point that germination commences.

The conidia of _P. densa_ are similar, but the germination is
different. When placed in a drop of water, under favourable
circumstances, the following changes may be observed in from four to
six hours. The protoplasm, at first uniformly distributed in all the
conidia, appears strewn with semi-lenticular, and nearly equidistant
vacuoles, of which the plane face is immediately in contact with the
periphery of the protoplasm. These vacuoles number from sixteen to
eighteen in _P. macrocarpa_, but are less numerous in _P. densa_. A
short time after the appearance of the vacuoles the entire conidium
extends itself so that the papilla disappears. Suddenly it reappears,
elongates itself, its attenuated membrane vanishes, and the protoplasm
is expelled by the narrow opening that remains in place of the
papilla. In normal cases the protoplasm remains united in a single
mass that shows a clear but very delicate outline. When it has reached
the front of the opening in the conidium, which is thus emptied, the
mass remains immovable. In _P. densa_ it is at first of a very
irregular form, but assumes by degrees a regular globose shape. This
is deprived of a distinct membrane, the vacuoles that disappeared in
the expulsion again become visible, but soon disappear for a second
time. The globule becomes surrounded with a membrane of cellulose, and
soon puts out from the point opposite to the opening of the conidium a
thick tube which grows in the same manner as the germ-tube of the
conidia in other species. Sometimes the expulsion of the protoplasm is
not completely accomplished; a portion of it remaining in