stigmas capitate. Achaenium (fruit) coriaceous, 3-
winged, cordato-oyoid, 1-celled; cell ending in an
elongated neck, similarly winged, interrupted near
the middle by a deep sinus, the base and apex rounded.
_ Seed erect, filling the cell, somewhat triangular,
stipitate, beaked, albuminous; albumen wanting in
the beak round the radicle. Embryo axile, straight,
clavate; radicle superior, about twice the length of
the cotyledons.—Erect, ramous shrubs, branches alternate.
Leaves coriaceous, entire, 2-stipuled, sometimes
fascicled. Flowers hermaphrodite, fascicled,
2-3 from each fascicle of leaves, pedicelled; pedicels
articulated below the middle, thickened at the apex,
vaginate with sheathing bracts at the base. Fruit
drooping.
I have taken the liberty of abbreviating this generic
character, which in the original is very long. The
most curious feature of the plant here represented is
found in the fruit, which, at the neck, receives a
twist, by which the upper half of the wing becomes
alternate with the lower, giving the fruit the appearance
of having six wings, three above and three
below. In another species, this peculiarity is wanting,
the fruit in it not being so contorted.
1809. Pteropyrum oliverh (J. and S.), leaves
fascicled, obovate or oblong, or spathulate, or somewhat
roundish, linear, or fla t: terminal wings of the
fruit almost concealed by the larger alternate lower
ones.
Scinde. I am indebted to Dr. Stocks for my
specimens of this curious plant.
As the genus has not yet found its way into general
systematic works on Botany, I have felt it necessary
to give the generic character. I t is given at
full length in Walper’s Annals of Botany, vol. 1st,
n. ssa m ’
1810. Rumex Nepaeensis (Spreng.), glabrous,
verticelsremote,many-flowered: fructiferous branches
nearly leafless: valves ovate, oblong, obtuse, reticu-
lately-veined; one of them obsoletely grain-bearing,
furnished at the base with subulate fimbri*, naked
towards the apex, the bristles shorter than the
breadth of the valve: leaves acute, somewhat waved,
the lower ones ovate, oblong, cordate at the base; the
radical ones oblong, subcordate; upper ones lanceolate
: stem very ramous, furrowed, tlnck.
Neilgherries, frequent, also on the Pulney Mountains,
but less common.
Begoniaceie.
This very curious order, consisting at present of 3
genera and about 160 species, has hitherto so completely
set the natural system of botany and its expounders at
defiance, so far as regards finding relationships is
concerned, that I think I may almost hazard the
assertion, that these are at the present moment about
f W f y as they were in 1789, when Jussieu
published his genera with the genus Begonia placed
among his “ plant* incerte sedis.” Since then many
attempts have been made to find a suitable location
m the natural series. De Candolle placed it between
c henopodiacea and Polygonaceee in which he has
been followed by several excellent Botanists. Link
looks to the TJmbelliferce for affinities; Martius to
near Campanidace<B; Meisner turns thence
to the Euphorbiacece, and thinks he has found the
most suitable station in their vicinity; Lindley in his
( I
Nixus suggested their affinity with Cucurbitacea, and
has been followed by Endlicher and Brongniart, the
former, however, with the remark that it is a difficult
order, not closely associating with any yet known,
and whose true affinities are questionable. Lindley,
in his Vegetable Kingdom, still adheres to this view,
and places the order in his Cucurbital alliance. This
I think by far the best station yet indicated, but still
the affinity appears so remote, that for the present I
am almost disposed to go so far as to say that it has
no really near affinity in the living flora of the earth,
and that we must seek its relationships among the
fossil remains of a former world.
Lindley in his character of the order assigns 4
sepals to the male, and 5 to the female flowers. This
must be received with some latitude, as the numbers
differ in different species. In regard to the seed,
they are said, to be without albumen, which, in those
I have examined, is not the case, they having a rather
large albumen in proportion to the size of the seed.
On the subject of affinities, Lindley’s views seem
at first sight very paradoxical, but may after all,
like many other paradoxes, prove very near the
truth. He says “ the relationship of Detassa is well
made out,” though it has a decidedly 1-celled ovary,
with parietal phrnent*. To this I demur. Again,
after stating that the main objection to the association
of Begoniads and Cucurbits in the same alliance
is the apparent difference of their placentation—axile
in the former, parietal in the latter—he thus proceeds
to show that the distinction is one of words, rather
than of essential structure. “ The ovary of such
Begoniads (some species of Diploclinium) consists of
three carpels whose dorsal suture is winged, and
whose margins turn inwards for a considerable distance,
each margin forming a plate or placenta over
which the ovules are arranged. This, with the exception
of the wing proceeding from the dorsal suture,
is the structure of Cucumis." To understand this it
is necessary to observe that the midribs of the carpel-
lary leaves of a Cucurbit are opposite the points of attachment
of the seed (see a transverse section of a
cucumber), and that the white line, extending from the
centre of the fruit to the seed, is not the partition between
2 cells, but is the two inflexed margins of the
same carpel as shown in Diploclinium; while the intermediate
triangular fleshy semeniferous portions are
simply modified forms of the thin partition shown to
exist between the cells in all the following transverse
sections of the ovary and fruit of Begoniads.
This view is certainly very ingenious, and is borne
out b y .. what we see in Rhododendron and some
Gesneracece, where similar marginal inflections of the
carpellary leaves exist. This view of the structure
of a Pepo, which at once and for ever overturns the
one which I formerly advanced, leaves no doubt,
when taken in connexion with the identity of form
of the stigmas and some points of habit, that Begoniads
and Cucurbits more nearly associate with each
other than either does with almost any other in the
series of ^ natural orders. There are still however
many points of difference between them, though it
must be admitted that a great step has been made
towards becoming acquainted with their time relationships.
On the subject of the parietal position
of the placentas of Cucurbits, I confess I am not yet
quite a convert to the doctrine, still less so after
being told that the difference between those of a
Pepo and a Begonia is one of words rather than of
)