4 ANNALS OI THE ROYAL BOTANIC GARDEN, CALCUTTA.
iire, from the first, united so as to foiTO a nuicellulai- ovary with numerous parietal
placentas, and of those species of Anona and Aberemoa and RolKnia where the pistils cover a
convex receptacle and become syncai'pous after fertilization, the whole of the members of
the family hav-e multiple and perfectly apocarpous fmits. Each pistil, as a rule, ripens
into an indehiscont one or several-seeded, often pulpy, berry with membranous, leathery or
occasionally woody or bony pericarp which is in some instances highly coloured. Truly
follicular fruits are found only in Anaxagorea ; but the elongated fruits of certain species
of Xylopia and Alphonsea dehisce in an ii-regulai- way. As a rule, the ripe carpels of
Anonacece are ovoid or globular ; more or less elongated pod-like fruits are, however, not
\mcommon. Some of the latter, e.g., those of many species of Unona and of some species
of Orophea, are not only much elongated, but are also constricted between the seeds.
Traces of constriction in more or less elongated fruits are also to be met with in other
genei-a. It is found, however, that, from the form of the fruit, characters of only
specific value can be taken. In many species of Anonacece the ovules in each ovary are
numerous ; in otliers they are reduced to two or even to one. In the latter case the
ovules are usually ascending, but in a few they are basilar and erect. In cases where
the ovules ai-e numerous, the direction of individuals varies with their position on the
placenta, being descending, horizontal or ascending, according as they are attached near
its apex, middle or base. The arrangement of the ovules in a multi-ovular placenta
in either a single or in double vertical row is really a matter of no importance ;
for it is regulated by the exigencies of space and pressure during the swelling of the
ovules after fecundation.
It should not be forgotten in connection with this character that originally every
placenta is double. The number of the ovules in a pistil, and their arrangement in cases
where they ai-e numerous, are not characters which are co-related to others in a way
which makes them of much real value to the systematist except in an artificial manner.
An illustration of the use of this character in the latter manner is found in the separation
of Ellipeia from Uoaria by the number of the ovules, the floral organs in these genera
being in other respects practically alike. As already mentioned, the seeds of this family
contain invariably a large quantity of ruminate albumen which, in some species, is
penetrated by membranous projections or septa proceeding from the integuments. The
embryo is always small, and, in some species, it is minute. The integuments of the seed
are usually thin and almost invariably smooth, but, when dry, there often cling to the
testa shrivelled masses of the pulp which in so many species fills the spaces between the
seeds; and such dried masses appear often to have been mistaken for an arillus. No
doubt, however, a small true arillus does occur in a few species. Observations on the
behaviour of the seeds of the family in germination have hitherto been confined to that
of certain species of Anona. On this subject I make the following extract from Sir
John Lubbock's recently published and most interesting, work "On Seedlings":—
"The seedlings of Anona are strong and vigorons, but present a singular anomaly, inasinucli as the
cotyledons are in some spedos torn from the asis during germination. The seeds of A. laurifolia split
longitudinally, the strong radicle strikes straight downwards, fixing the seedling firmly in the soil, and the
hypocotyl forms an ascending or elongating loop, The large seed remains under ground, while the hypoootyl
at length pulls out the plumule, and straightens, leaving the cotyledons behind. If a seed is cut open at
this point the cotyledons are seen to be oblong or oval, obtuse, petiolate, almost equalling the length and
breadth of the seed, and much undulated or wrinkled in order to accommodate themselves to the inequalities
of the much ruminated endosperm. They exhibit an ascending, incurved, feather-nerved venation very
AN ONAGER OF BRITISH I^'DIA.
similar to that of the leaves, and at this stage aro surrounded or included between two equal halves of
the - endosperm with a small quantity of the same between them. A. cherimoUa behaves io exactly the same
-way, and the cotyledons are invariably torn from the axis and left in the subterranean seed."
The inflorescence in this family presents no character of note with the exception of
the curious fasciation of the peduncles in the species gi-ouped under the genus Artaboirys.
Few-flowered, axillary or leaf-opposed cymes or racemes, and even solitary flowers, prevail
in the order; panicles being rare, and spikes unknown. The simple alternate ex-stipulate
leaves of the family are always cntii-e and, for the most pai-t, tbey are membranous or
sub-coriaceous; rarely coriaceous, and never fleshy. Glandular dots or markings occur in
the leaves of a few species—notably in those of some species of Anona.
Although the genera Ancna, Uvaria and Xylopia arc Linna^an, and Unona is of
Linnfeus filius, the earliest recognition of these plants as the type of a family is to
be found at page 365 of the second volume of Adanson's Families des Fiantes (published
in 1763), where that author distinguished the group which he called Los Anones. Les
Anones of Adanson, however, included various plants which have in more recent times been
relegated to other families than Anonacece. The whole of the then-known plants which
have since been foi-med into the families Magnoliacea; and Mcnispermacece, ccrtain of
the Dilleniacece and Ranunculaeece of the present day, together with the genera Oehna
and Fagara, were grouped by Adanson with the true Anonaceas as we now understand
them. The systematic writer who next gave special attention to the family was
Antoine de Jussieu; and his Anonecc^ in addition to the Linoiean genera ah-eady
mentioned, included the genus Cananga which had been founded by Aublet in 1775.
Jussieu, moreover, referred the majority of the other genera included by Adanson to
the natural order Magnolice. The inventor of the name Anonacece was L. C. Richard;
but the first botanist who really understood the group and who defined the limits of
it pretty much as they stand to-day was Dunal, whose Monograph appeared in 1817.
That author included in the family nine genera, viz., the Linnasan Anona, Uvaria and
Xylopia, Unona of the younger Linnreus (in which he included Desmos and Melodonm
of Loureiro), Forcelia (founded by Ruia and Pavon in 1794), Guattcria, Ruiz and Pavon
(also 1794, but now admitted to be reducible to the earlier published Cananga of
Aublet), Asimina of Adanson, and his own genus Monoclora (= Anona Myristica, Gasrtn.)
The only extraneous genus which Dunal admitted was Kadsura. Dunal's arrangement
was followed by M. Alph. DeCandolle in his Memoire sur la Familie des Anonaeees
published in 1832, except that he excluded Kadsura, himself described fom- new genera
(viz., Hahzelia, Cceloeline., Miliusa, and Ilexalobus), and included the genera liollinia, AnaX'
agorea, Artaboirys, Orophea, Bocagea, Fohjalthia, and Duguetia which had been published
since Dunal wrote. Asimina and Foreelia, however, he reduced to Voaria. Still later Blume,
in his Bißragen and Flora Javas, either founded as genera, or separated as sections
which have since been maintained as genera, Mitrephora, Stclechocarpus, Goniothalamus and
Oxymitra. And it was Blume who first recognized the true affinities of KaJsura, and
who erected it, with its near ally Schizandra, into a new family named Schizandrece
a group which was united to Magnoliacece by Drs. Hooker and Thomson in 1862, and
which has by most subsequent authors been treated as a tribe of that family.
As DeCandolle left it in 1832, the family of Anonaeeas stood very much as it
does to-day, although its members were far less numerous. It consisted for the most
part of plants of which the sidient features were the ternary gymmtery of the