Generitype: A. rujcsccns Aubl.
We have reexamined fruits of 23 of the 30 currently recognized species of Abuta,
of all, that is, except J. chiapasensis, A. convexa, A. mycetandra, A. negroensis, A.
rufescens, A. steyermarkii, and A. vaupesensis. The sample is believed to give a
reasonably complete view of the drupe in this, the largest genus of Menispermaceae
in the Americas.
A. antioquiana: Hno Daniel 3355 (type coll.).
A. aristeguietae: Steyermark 97482 (type coll.), Bruno J. Manara s n (4/5 Jun 1966), Woytkowski 5541.
A. barbata: Florshiitz & Maas 2448, Abraham 315.
A. brevifolia: Black 49-8156, Ducke s n (Mus. Par. Emilio Goeldi 8976).
A. bullata: Gleason 358 (type coll.).
A. candollei: Boschwezen 6747, coll. undcsign. s n ("E", 16/VII-1939).
A. colombiana: Cuatrecasas 14028.
A. dwyerana; Duke 9425 (type coll.), Dwyer et al. 7250.
A. grandifolia: Krukoff 7812, Prance et al. 4068.
A. grisebachii: Krukoff 8660, Froes 21277
A. imene: Froes 21142, 26329, Krukoff 7185, 8605, Pires 239, Holt & Blake 685, 690, Garcia-Barriga 13935.
A. longa: Steyermark 90079 (type coll.).
A. obovata: de la Cruz 2250 (type coll.), Maguire & Politi 28287, 28442, Maguire et al. 29403, Pires et al. 50699, For. Dept. (Guiana) 4233, 5365, Steyermark 60395.
A. pahni: Cardona 2601, Krukoff 5437, Fuller 98 (type coll. A. ecuadorensis).
A. panamensis: Molina et al. 17328, Brenes 20501.
A. panurensis: INPA 3774; Oliveira 94.
A. racemosa: Romero-Castaheda 4877.
A. sandwithiana: Irwin et al. 48289, Krukoff 7051, Prance et al. 6229, 6635.
A. seemanni: Cuatrecasas 17672, 17709.
A. selloana: Mello Barreto 1662, 1777, Reitz & Klein 1868, 6562.
A. solimoesensis: Froes 23964 (type coll.), Schunke 2243.
A. splendida: Krukoff 8030, 11083 (type coll.), Buchtien 1901, Froes 21014, 22209,Steyermark 56701, Steyermark & Rabe 96553.
A. velutina: Froes 12001/20, Pires 1121, Tate 959 (type coll.), Zavortink 2280, 2281.
In its exterior form the dried fruit of Abuta resembles that of Curarea, being
obovoid or oblong-obovoid, a trifle compressed laterally, with an induplicate stone
bringing the style-scar down to the level of the fruit-stalk. The vertical transverse
condyle is that of Curarea, but the embryo is that of tribe Anomospermeae, linearvermiform
and embedded in laminated endosperm, each plate of which is clothed in
a membranous integument. Apart from expectable variation (both intra- and interspecific)
in thickness of the mealy-coriaceous exocarp, in development of a pulpy
mesocarp (never thick), and thickness of the shell, the drupe is extremely uniform in
basic design. Externally the stone has a pattern of one dorsal and on each side one
lateral groove extending the length of its long axis, these being connected by a network
of smaller and finer channels. Within, the testa is smooth and lustrous.
The pericarp in most Abuta fruits is dry and firm, closely investing the stone,
but in several there are traces of pulpy mesocarp and in a few, notably A. imene, A.
sandwithiana, and J. solimoesensis, the pulp forms a considerable layer which
evaporates in drying, so that the exocarp collapses and becomes wrinkled or ruptured
in the press. A few species (A. dwyerana, A. aristeguietae) are noteworthy for the
thickness of the firm pericarp which is dotted externally with Httle pallid lenticels.
The species of Abuta can for the most part be grouped into two series by thickness of
the endocarp as measured in section. Thin-shelled species (testa 0.2-0.5 mm thick)
are in the majority; thick-shelled (testa 0.6-1.5 mm thick) are A. antioquiana, A.
aristeguietae, A. bullata, A. longa, and A. seemanni. A few, notably A. splendida and
A. obovata, are known to have thick- and thin-shelled variants, and this phenomenon may turn out to be commoner than is knovm at present. The thickest shell seen is that of A. longa, in which the incised sculpture extends 1 m m deep into a wall up to 1.5 mm thick. Even if we had knowledge of drupes of all species it seems unlikely that a workable key could be based on them, although their characters are certainly valuable at the specific level.
From Anomaspermum and Orthomene the genus Abuta differs in the curvature and emplacement of the seed on a vertical condyle entering from the base and in the apetalous flowers. Caryomene has a similar, inversely U-shaped seed, but differs in its internally sculptured stone of extremely thick woody texture and in the absence of membranous integument around each plate of endosperm. The flower of Caryomene is still unknown, so that the difference in that direction cannot be evaluated. From Telitoxicum, perhaps its nearest kindred, .Abuta differs hardly at all in characters of the drupe but is in practice easily distinguished by its pilnerved, not penninerved, leaf-blade.s and its apetalous staminate flower.
While knowledge of Abuta has increased substantially in recent years, there are still many vexatious gaps in it, some of them serious. Two species, one of them, generitype, are known only from the leaves. Of nine species staminate flowers are not known, and of seven we have no fruits. Several abutas of which we claim to have knowledge of the flower or the fruit are known in reality from only one fertile collection. Of most species we know nothing or next to nothing about intraspecific variation in size of flower-parts, in dilation or concretion of the filaments, or size of the drupe. In practice, therefore, far too many taxa in the genus must be recognized and defined in terms of foliage characters, often in themselves distinctive enough, but inherently unsatisfactory from the systematic viewpoint. Moreover leaves of a given species can vary greatly, according to age and position on the vine, in shape, texture, and density of pubescence, shade leaves being generally larger and thinner-textured than sun-leaves associated with flowers in the forest canopy.
The acquisition of flowering material of a particular sex of a particular species of Abuta being as much a matter of chance as of will or effort, we foresee a long term of years before the taxonomy of the genus can be placed on a really firm footing. Certainly we lack the data necessary for recognition of subgeneric groups. Eichler in Flora Brasiliensis arranged the few species then known into four sections, defined by androecial characters, but subsequent discoveries have not fitted at all neatly into the preconceived categories. The androecium of Abuta is remarkably polymorphic, due entirely, we think, to modifications of the filaments. These may be either free or partly united into a column, and there is a tendency for the three outer ones to dwindle into staminodes. Due to differential growth of the distal end and connective, the anther-sacs may be displaced so as to He collaterally extrorse, collaterally introrse, or collaterally terminal, or may be separated both dorsally and ventrally by connective tissue and then simply lateral. The dehiscence is normally by vertical slits, but due to various displacements of the sac, whether around the head of the filament or by forward tilting of its tip, the slit may come to appear oblique or transverse. In A. grandifolia the anther-sac is constricted at the middle, the whole anther appearing 4-lobed ("tetracoccous") in consequence. In a few species the filament itself is hispidulous, but in the majority quite hairless.
Androecial diversity is one of the most interesting features of Abuta and
deserves attentive study as opportunity arises. For the present we cannot find any
positive correlation between the types of filament and other macroscopic characters,
and therefore refrain from any attempt to elaborate or correct the Eichlerian concept
of sections. In its place, as a temporary expedient dictated by our ignorance, we
have arranged the species into four groups (three of them loosely equivalent to sections of Eichler) which we suppose from a general and gross similarity of aspect to contain genuinely related species: