ous form; filaments free; anthers normal;
carpel transformed into a true leaf with a long stalk provided
at the base, with two stipules, terminal leaflet, solitary,
green, with no trace of ovules. Sometimes a second carpellary
leaf, similar to the first, is formed; in other cases the
central axis of the flower is occasionally prolonged into a
head of young flowers--median prolification. In some few
instances the calyx is not at all altered, but the carpellary
leaf is trifoliolate, or even quinquefoliolate, the corolla
being then absent. The heads of flowers in this first form have
the aspect of little tufts of leaves.

2. Each of the teeth of the calyx is represented by a long
stalk, terminated by a single articulated leaflet, the
bi-labiate form of the calyx is still recognisable; the two
upper petals are united, the three lower separate; the tube of
the calyx is not deformed and seems to be formed of the
petioles of the sepals united by their stipules. In this second
class of cases the corolla is papilionaceous, the filaments
free, the carpellary leaf on a long stalk provided with
stipules, its blade more or less like the usual carpel, with
its margins disunited or more commonly united with the ovules
in the interior, sometimes represented by a foliaceous, dentate
primine only. In one case the carpel was closed above, gaping
below, where it gave origin to several leaflets, the lower ones
oval, dentate, like ordinary leaflets, the upper ones merely
lanceolate, leafy lobes, representing the primine reduced to a
foliaceous condition. Inflorescence--a head with leafy flowers
on long stalks, which are longer at the circumference than in
the centre.

3. Calyx-teeth lance-shaped, acuminate; corolla more or less
regular, arrested in its development and scarcely exceeding the
tube of the calyx within which it is crumpled up; stamens but
little changed; carpellary leaf on a short stalk, not exceeding
the calyx tube, but the ovarian portion very long, and provided
with abortive ovules.

These three groups will be found to include most of the forms
under which frondescence of the clover blossoms occurs, but
there are, of course, intermediate forms not readily to be
grouped under either of the above heads. Such are the cases
brought u