Menispermum abuta Lamarck,Encyc. A: 100. 1796.
Abuta scandens Barrere, De Candolle Prodr. 1: 103, synonym. 1824.
Menispermum abuta L., Descourt Fl. Ant. 4: 101. PI. 256. 1827.
Cocculus abuta Kosteletsky, Allg. Med-pharm. Fl. 2 501. 1833.
We are citing here all collections of this species examined by us in connection with this paper.
Staminate flowers and fruits of this species are not known.
Pinkley states on the label of Pinkley 286: "used in Curare of Kofan Indians";
the label on Schultes & Cabrera 16569 reads: "this is one of the components of
Yacuna arrow poison."
The concept of an Abuta rufescens dispersed almost throughout intertropical
South America east of the Andes, and north interruptedly to Costa Rica was formulated
in the monograph (Brittonia 3: 66) and expanded in the supplements until
summarized in our last paper [Mem. N. Y. Bot. Gard. 20(2): 13] ; it is here substantially
modified. The inclusive deUmitation was long maintained for two reasons:
the material (by now extensive, over 80 collections having been now revised) is
largely sterile, or if fertile then pistillate; and no pragmatic method of sorting specimens
by means of vegetative characters had been developed. Having now reexamined
all available fruiting and staminate flowering collections of the complex we find
that important differences in the androecium correspond with detectable reticulationpatterns
in the leaf-blade, and, further, that the categories so established meet the
test of phytogeographic probability, each type filling a well-known area of endemism,
larger or smaller.
What we may call for purposes of discussion the rufescens complex is defined
chiefly by foliage. The leaf-blade is relatively ample, 5-plinerved, often (but not
invariably) bullate, with primary and scalariform secondary venation usually impressed
above and elevated beneath. At maturity the blade is glabrous above (except
for rare remnants of tomentum along the depressed primary veins) but remains
permanently hirsutulous, pilosulous, or velvety-tomentose beneath with soft, erect,
ascending, or sometimes curly and entangled hairs, commonly (at least when dry)
of a brownish or sordid yellowish color. For the purpose of this note we disregard
the well-known A. grisebachii and A. candollei, species which would fall into the
rufescens complex if it were defined purely and simply in leaf-terms, for these
are known to differ greatly in staminate flower and can be recognized in practice
by facies alone; but we include A. splendida, in which the dorsal leaf-surface is
normally densely white-pannose-tomentulose, and will mention in passing A. barbata,
A. aristeguietae, and the Central American A. steyermarkii, the foliage of which
would pass easily for that of A. rufescens sens. lat.
Our narrower definition of A. rufescens includes only those plants in which the
leathery leaf-blade is perfectly smooth on the upper face, the fine reticulation being
completely immersed and concealed beneath the epidermis, and on the lower side
softly pilosulous with erect brownish hairs closely spaced along the elevated veins
and veinlets but not so closely as to conceal the surface of the moderately depressed
areoles. It is an extraordinary fact that although foliage of this type was collected
two hundred years ago by Aublet and has been encountered since at points widely
scattered through middle Amazonian Brazil, we still lack flowers of either sex, and
the fruit is known only from Aublet's figure, and is not surely relevant. The only
other member of the complex in which the leaf-blade lacks visible reticulation on
the upper side is A. splendida, the known range of which coincides with (and extends
beyond) that of A. rufescens sens. str. The leaf of A. splendida is quite similar to
that of A. rufescens in texture and venation but is on the average somewhat smaller
and densely pannose-tomentose beneath with ordinarily whitish hairs. It seems
not unlikely that our concept of A. rufescens is based on a selection of shade-leaves
and leaves from young vines which, if and when they can reach the forest canopy,
might bear the more densely tomentose foliage and the flowers of A. splendida. At
present we do not have any evidence of organic connection between the two series of
specimens and therefore we refrain from reducing A. splendida to synonymy of A.
rufescens. The problem requires field-observation for satisfactory solution.
Typification of A. rufescens deserves comment. Paradoxically the difficulties in interpretation of the protologue are due not, as so often happens, to lack of data but to a plethora of conflicting data and specimens. There are sheets from Aublet's herbarium at Geneva, at Paris, and at the British Museum, but these, all sterile, contain elements of three species: the A. rufescens of this note, A. candollei Tr. & PI., and of Curarea candicans. The mixture was noted very early, by Planchon, and the specimen which furnished data for the account of A. rufescens in de Candolle's Prodromus thereupon became the type of .A. candollei Tr. & PI. Another, at Paris, which according to the label was given by Aublet himself to the elder Jussieu as A. rufescens, represents the same species. A second sheet at Paris contains the misplaced Curarea element; and only the British Museum sheet, which, following Planchon, Miers, and Diels, all of whom examined it, we acknowledge as lectotype of A. rufescens, represents the species under consideration here. This lectotypification, effected by Triana & Planchon (Ann. Sci. Nat. IV. 17: 47), is nevertheless probably contrary to the spirit of the protologue. Aublet's account of A. rufescens obviously covers more than one entity, for he specifically mentions two types of Pareira Brava, the white and the red, both of which furnished, by infusion of the vines, a specific against disorders of the liver and gallstones, and was for this purpose exported to Europe from Cayenne (Note: not the Pareira Brava of southern Brazil, which is Chondrodendron). The principal description of A. rufescens refers to the White Pareira Brava, which Aublet had observed both on the island of Cayenne and in "almost all the forests of Guiana" that he had visited. It seems entirely probable that this was in reality Abuta candollei, which is known to be common on the coastal lowlands of French Guiana. The Red Pareira Brava, said to differ from the White in having the leaves beneath and young branchlets covered with "duvet roussatre," not "cendre," is treated (despite the epithet rufescens applied to the whole) only as a minor variety. We cannot escape the conclusion that only this form corresponds to the A. rufescens of Triana and the British Museum specimens, and that the intention of Aublet has been altered in consequence. Aublet says of the Red Pareira Brava that it grew in the same places as the White, both on Cayenne and the mainland, but this raises another difficulty, for the A. rufescens represented by Aublet's specimens at BM has never again been encountered, so far as we are aware, in French Guiana. We can only guess at the origin of these leaves which match so very closely in the last detail the material of Brazihan Amazonia which embodies our concept today of Abuta rufescens Aubl. sensu Triana & Planchon. We notice, however, concealed in the protologue of A. rufescens, a possible, indeed plausible explanation of the appearance in a collection of plants from French Guiana of leaves certainly referred to an Amazonian species. O n leaving He de France in 1762, Aublet left growing in the garden of the fort a flourishing plant of Pareira Brava, brought to him from Brazil by a Father Series. We suppose it possible that the leaves of Abuta candollei that reached the herbaria of Jussieu and de Candolle were collected in the wild and that the leaves retained in Aublet's own herbarium, now at British Museum, were taken (what more likely?) from the plant he had himself cultivated at Cayenne. Leaves shown in Aublet's plate clearly were drawn from the species that we here call A. rufescens. If our suggestion is correct, we would still have to explain the existence in French Guiana of a second .Abuta, other than white-tomentulose A. candollei, that might correspond with the native Red Pareira Brava that Aublet distinguished by its rufous indument. A candidate neatly tailored for this role is .Abuta barbata Miers, present in and indeed actually described from Cayenne. Triana & Planchon already supposed that the plant collected by Sagot which became the type of A. barbata was the very same as that called A. rufescens by Aublet; not, of course, the plant represented in Herb. Aublet. at BM, but that which Aublet had seen growing in Guiana and called Pareira Brava Rouge.
The plate accompanying Aublet's account of Abuta rufescens shows leaves and
fruits. The author tells that although he found the species common in the forests of
Guiana he never encountered a vine in flower and only once met with the fruit, which
was seen ripe in January in a locality recorded as the "quartier de Caux." W e guess,
but cannot be sure, that this m a y be the settlement at the mouth of the Approuague
River about 55 k m southeast of Cayenne that is spelled K a w on some modern maps.
It would be interesting to know what species of .Abuta is found there, for the drupes
seem not to have survived in the Aublet herbarium and as they are shown detached
from the leaves they may weU not belong to the same plant as the leaves. It is perhaps
enough to regard the fruiting element in Aublet's plate as certainly representing the
genus .Abuta, the pubUcation of which dates from this plate. The sections of the
drupe, showing the emplacement of the seed and the elaborate folding of the endosperm,
are characteristic of the genus.
As already noted, A. rufescens and A. splendida (assuming they are distinct) are the only members of the present complex characterized by leaf-blades smooth above; but we have seen two rather striking collections which while sharing this feature differ in that the lower leaf-surface is only thinly spreading-pilosulous and has much less deeply excavated areoles also. Both collections are from relatively high ground on tributaries of Rio Negro (base of Auyantepui, Alto Caroni, 800 m, in Bolivar, Venezuela, fr, Cardona 2601; summit of Curicuriari, Amazonas, Brazil, sterile, Schultes & Lopez 9827a, both NY), and possibly represent a related but undescribed species. However, we refrain from naming it until staminate flowers are forthcoming.
Returning to the dismemberment of A. rufescens of recent literature, we note that all the material now excluded from the species has leaves at least faintly reticulate above, but is diverse in other ways. The Brazilian material is divided between: A. convexa, of which the staminate flower and androecium closely resemble those of
A. splendida; A. pahni in which relatively thin and short pubescence of the leaves is
combined with a completely different flower; and a new, imperfectly known A.
mycetandra. From the Andean foothills, Venezuela, and central America we recognize
A. aristeguietae and .1. steyermarkii. In foliage and pubescence the first of these
closely resembles .A. pahni. but differs in its large drupe; no staminate flower is known.
The foUage of A. steyermarkii resembles that of A. convexa, but the very large
subsessile pistillate flowers are quite distinctive; here it is lack of both fruit and
staminate flower that prevents a thorough separation of what we do not doubt to
be a perfectly distinct species.
It remains only to mention that all classic accounts of A. rufescens have been
contaminated by elements of one or more of the foregoing segregates. Eichler's
superb plate of A. rufescens in Flora Brasiliensis 13(1): tab. 39 portrays A. convexa;
and the A. rufescens of Diels (1910) consisted entirely of A. convexa except for the
sterile types of A. rufescens itself and of A. pahni.
The stems of this species are round and not flat (Mathias & Taylor 3553, and
Heckel states: "le pareira blanc des Creoles, WAbuta rufescens d'Aublet, sert a
faire des tisanes avec sa tige et ses racines, contre les obstructions du foie et la morsure
des serpents. On pretend que la plante entre dans la composition de certain curares.
On en applique les feuilles sur les ulceres." (Heckel, Edouard, Les plantes medicinales
et toxiques de la Guyane Frangaise, pp. 1-93. 1897.) We were unable to ascertain as
yet whether this statement refers to Abuta rufescens, Curarea candicans, or Abuta