Russel Wallace : Alfred Russell Wallace (sic)
There seems to be sufficient variety of form and structure to justify the establishment of five genera.
1. The true Gymnotus (of which the Gymnotus electricus appears to be the only well-known species), characterized by the anal fin reaching the extremity of the tail, which is flattened; by the air-bladder extending almost the entire length, in a cavity beneath the vertebræ; and by having a single row of short acute teeth in each jaw.
2. The genus Carapus, to which five of my species belong. These have the tail cylindrical and pointed, extending beyond the anal fin; a band of minute teeth in each jaw; and a double air-bladder, generally of very small size. One of my species appears to be identical with Carapus brachyurus of Bloch.
3. A form, of which I have but one representative, which has a deep body, a rather large mouth, but no teeth, and a small round single air-bladder.
4. Two long-jawed species, which have a very small mouth, no teeth, and no air-bladder. The larger of these is probably the Gymnotus rostratus of Schneider.
5. The genus Apteronotus, which differs from all the preceding in having a small, but perfectly-formed and rayed caudal fin, a rather large mouth, with the lower jaw shutting within the upper, and the teeth rather acute and prominent in a row on the sides of the jaws only. My representative of this genus appears to be quite distinct from Apteronotus albifrons of Lacépède.
These fishes were all found near the sources of the Rio Negro and [[p. 76]] Orinoco, one of the most central positions in South America. They are most abundant in the smaller streams, and feed on minute aquatic insects. None of them, except the common Gymnotus, have any electrical properties. They are all eaten, though, owing to the number of forked or branched bones in every part of their bodies, they are not much esteemed.
The situation of the vent in these fishes is very peculiar, the intestine passing forwards from the stomach, instead of backwards, as is usually the case, so that they have the anus situated under the throat; in one of the long-snouted species it is actually considerably in front of the eyes, a peculiarity which I believe does not occur in any other vertebrated animal.
This fish, too, is remarkable for the very singular manner in which it is said to feed. It is asserted that it lives principally on ants and white ants, which it obtains by laying its tail out upon land. The ants, attracted by its mucous covering, crawl thickly upon it, when the fish dives down and leaves the ants struggling upon the surface of the water, where it is enabled to eat them at its leisure. The Indians assert that, when fishing at night, they often see this take place.
To give some idea of the distribution of fishes in the rivers of South America, I may mention, that of 205 species which I found in the Rio Negro,--
43 were spinous-finned fishes, principally Percidæ and Labridæ;
Around the middle of the 19th Century the immense diversity of the South American freshwater fish fauna was only just beginning to be revealed, with the early 1800s publications of Lacépède, Bloch & Schneider, Cuvier, and Spix & Agassiz, and the Amazonian collections made by J. B. Spix & K. F. P. von Martius (1817-1820), and Johann Natterer (1817-1835). During the second half of his four year (1848-1852) expedition to the Amazon basin spent in the Rio Negro and Rio Uaupés, Wallace developed an increasing interest for the ichthyofauna of that region which resulted in a collection of almost 450 specimens, plus 212 pencil drawings of fishes, with notes on most of them. Had the specimens not been destroyed, their study would certainly have resulted in the description of many new species unknown at that time. The scientific value of the material that Wallace was planning to bring back to England may be appreciated through the collection of his fish drawings which I have recently (2002) arranged and published as the monograph Peixes do Rio Negro / Fishes of the Rio Negro.
The ten species of Neotropical electric fishes which Wallace refers to in this 1853 paper are all pictured in his collection of drawings, and at least four of them were undescribed at that time (Wallace's genera 3, 4 & 5, and part of 2). Wallace assembled the ten into a five group arrangement bearing some similarity to the current family level classification of the Order Gymnotiformes. Genera 1 and 5 represent the Gymnotidae and Apteronotidae, respectively (although the former species is sometimes placed in its own family, the Electrophoridae). Genera 3 and 4 are proposed as new by Wallace and represent the Hypopomidae and Rhamphicthyidae, respectively. Genus 2 includes representatives of the Gymnotidae, Hypopomidae and Sternopygidae.
Contrary to what Wallace says, not only the Electric eel, Electrophorus electricus (referred to by Wallace as Gymnotus electricus), has electrical properties. Today it is known that all living gymnotiforms are capable of producing and detecting weak electric fields. Electrophorus differs in its ability to produce strong electric discharges that stun prey or ward off predators.
An updated estimate of the fish diversity surveyed by Wallace in the Rio Negro gives a total of approximately 180 species distributed in 33 families and 110 genera--a significant figure considering he was a non-specialist and was involved in so many other activities at the same time. A 1988 study gives an estimate of 450 fish species for the entire Rio Negro drainage.