Generitype: S. cayennensis Bentham.
This genus resembles Chondrodendron in its pluricyclate staminate perianth and
six (in one species 9-15) carpels, but differs in the great development of the carpophores,
which, anthesis past, carry the forming drupe aloft on a slender stalk, often
mistaken for a pedicel. The carpophores vary considerably in length, reaching
exceptionally 1.7 cm in S. brachypoda; they are connate at base into a common
column, grooved lengthwise (when dry), and silky- or velvety-puberulent. In the
structure of the carpophore Sciadotenia resembles Curarea except that in the latter
there are only three carpels and the three carpophores that develop under them after
fertilization are free from each other, radiating from a common point. Curarea
differs further in the simpler perianth, with only two cycles of sepals. Curarea
and Chondrodendron share the striking character of a minutely tomentellous indument
on the back of the leaves. The majority of Sciadotenia have glabrous or thinly
strigulose foliage, and the exceptions, S. paracusis and S. sagotiana, have leaves
velvety-tomentellous with relatively coarse entangled hairs. In the form and texture
of the endocarp, Sciadotenia hardly differs from Chondrodendron and has the condyle
and embryo common to all American Triclisieae. The endocarp of Curarea is
papery and lacks external sculpture. Leaves of Sciadotenia are initially plinerved,
and the primary veins are connected by transverse secondaries running more or less
horizontally between them in a scalariform pattern. In the context of the tribe
this is a helpful character in sorting sterile material, but similar venation patterns
occur also in some Abuta. The genus is defined principally by the umbellate cluster
of gynophores, the character which is commemorated in the generic name.
The genus is poorly collected, and its history has from the first been clouded by
lack of securely matched staminate inflorescences and fruits. In connection with this
study we have seen staminate flowering material of aU but two described species,
mature drupes of ten, immature drupes of two, and the pistillate inflorescence
with carpophores but no drupes of two. Three species are known to us only from
staminate material, two from pistillate only. W e Hst herewith the vouchers actually
studied in preparation of this paper, the existence of a matching staminate specimen
being noted in the column at right:
Two types of pistillate flower are found in Sciadotenia. That of S. cayennensis, borne singly on a long, flexuously pendulous stalk, has 9-15 ovaries; the endocarp is Innately incurved through about three fourths of an arc over an umbonate condyle so that the style-scar is finally situated about halfway between the insertion of the carpophore and top of the drupe. In all other species the pistillate flowers are, if solitary, borne on a much shorter peduncle; more often there are two or more arranged in a short raceme on a stiff, either erect or recurved axis. In any case all have exactly six ovaries giving rise to hippocrepiform endocarps folded over a vertical, septiform condyle, the style-scar brought down almost to the point of insertion. The embryo of S. cayennensis consists of two half-cyUndrical cotyledons that remain plump, smooth, and lustrous; in other species the embryo is usually wrinkled and dull externally. Except for the slightly anomalous character of the embryo in S. cayennensis, the fruit of Sciadotenia is rather uniform in character, the endocarp varying somewhat in depth of sculpture depending on the thickness of its leathery or crustaceous testa, and to minor degree in size and outline. Drupes of some species have thin mucilaginous mesocarp so that the exocarp closely invests the stone and becomes little wrinkled in drying. In others there is a copious pulp, red or orange when ripe, which collapses in drying, leaving a much wrinkled skin around the stone. Little is known, however, about the stages of ripening, and the differences in succulence of the fruit, as now seen in specimens, m a y be more apparent than real. In general the drupe offers Httle in the way of differential characters, and helps little to illuminate relationships within the genus.
relatively well-known staminate inflorescence offers much of systematic
interest and importance. In the majority of species the flower is rather uniform,
consisting of a little cone of graduated, submembranous, externally pilosulous or
loosely strigulose sepals that suggests the involucre of some Compositae. The valvate
inner cycle of sepals is most commonly a little shorter than the cycle next below
it or, if longer, only a trifle so. Within the valvate sepals are six abruptly smaller,
obovate or oblanceolate (but not clawed), slightly concave petals of membranous
texture, pubescent on the back; and within these six stamens, sometimes free,
sometimes partly adherent into a synandrium, bearing at apex a pair of terminal
anther-sacs contiguous at least introrsely and opening by oblique, confluent slits.
This type of flower and stamen is well illustrated in Flora Brasiliensis 13(1): tab.
49, figs. I, II, and has been considered, since the monograph of Diels, as typical for
the genus. In reality, however, it is not the flower of the generitype, S. cayennensis,
nor that of two species, S. toxifera and S. solimoesana, described in later years.
The three species just mentioned are alike and differ from all the rest in having
broad, ovate sepals of firm texture, silvery outside with a very short and closely
appressed strigulose pubescence, and strongly graduated, the valvate inner cycle
being nearly twice as long as the cycle next below it. Associated with this outer
perianth are distinctly clawed petals dilated upward into a blade either membranous
or fleshy-thickened (as will be brought out below); and a set of six stamens, again
either free or partly fused, greatly enlarged at the incurved tip so as to bear the
anther-sacs laterally, widely separated both back and front by a broad connective
produced beyond them into a conic point.
The petal-blades of S. cayennensis are membranous, unappendaged, subcordate
at base; the filaments are free above the extreme base; and as already mentioned,
the pistillate flower is at once solitary and long-stalked, has more than six ovaries,
and gives rise to a seed substantially different from that of all congeners. The species
is morphologically isolated in the genus, and we propose for it a mono typical sect.
The staminate flower of S. toxifera and S. solimoesana is essentially that of sect.
Sciadotenia in the texture, pubescence, and strong graduation of the sepals, but the
petals bear on the inner face, above the claw, two callous ridges, in S. toxifera tilted
inward distally so as to form an inverted V, in S. solimoesana vertical and parallel.
With this perianth is associated a hexandrous synandrium, an erect, few-flowered
pistillate inflorescence, and a horseshoe seed folded on a septiform condyle. When
we first encountered an example of this type of staminate flower, so different from
anything hitherto supposed to occur in Sciadotenia, we were misled into thinking
it generically distinct, and we based the genus Tylopetalum upon it. When in due
course we came to review Sciadotenia, we recognized instantly in recently collected
flowering specimens of S. toxifera (Schunke 1969-144) the peculiar Httle flower
of Tylopetalum abutiforme. Now that Tylopetalum is known to have essentially
the Sciadotenia-like pistillate organization of S. toxifera, we can no longer evaluate
it as a distinct genus although within Sciadotenia it forms a strongly marked group,
to which we accord a sectional status equivalent to that of sect. Sciadotenia.
The remaining 15 species of Sciadotenia, uniform in essential characters of the staminate flower, fall possibly into two groups, one with pistillate inflorescence more or less 1-3-flowered, and shortly pedunculate; the other with a stiffly erect pistillate axis bearing, like that of sect. Tylopetalum, a few-flowered spike of more or less erect drupes. The first type is illustrated by Eichler [Fl. Bras. 13(1): tab. 47, fig. 3] as S. amazonica. The second is typified by S. paracusis, the generitype of Sychnosepalum Eichl. So far as we can tell at present, the drupes associated with these two inflorescence-types are alike, just as the staminate flowers are alike, so that they may best be treated as forming a single somewhat diverse group. For this we adopt at sectional level the name Sychnosepalum Eichl., descriptive of the many leaved, cone-Hke staminate perianth.
The differential characters of the three sections of Sciadotenia are summarized in the key following. Two nomenclatural transfers are required: