OSTARIOPHYSI 



Characins, Gymnotids, Carps and Catfishes 



As IS well known, this immense assemblage of some five thousand species, mostly of 

 freshwater fishes, constituting the order Ostariophysi, exhibits the utmost diversity of 

 habitus in body- and skull-form, but is shown to be a natural group by the common 

 possession of the elaborate Weberian apparatus connecting the swim-bladder with the inner 

 ear, which is essentially the same in all the families of the order. These ossicles, which are 

 believed to be derived from the ribs and neural arches of the four anterior vertebrae, trans- 

 mit vibrations to the membraneous labyrinth of the ear and probably serve to increase the 

 sense of hearing (Tate Regan, 1929, p. 315). It seems curious that the importance and 

 significance of this unique apparatus seem to have been greatly underestimated by Garstang 

 (1932, p. 241), who refers depreciatively to the "one peculiar specialization of a few anterior 

 vertebrae' as being inadequate for excluding the Ostariophysi from a place "on the top 

 shelf" of the Teleostei, along with the Isospondyli. But it is doubtful whether there is 

 any taxonomic and phylogenetic group in the whole series of vertebrates that is more 

 deeply stamped than is the Ostariophysi with the mark of unitary origin through the 

 possession of an elaborate "basic patent" and basic pattern. If this is not a strictly 

 circumscribed natural group, where in all the animal kingdom shall one be found.' And 

 how is it possible to speak of "one peculiar specialization of a few anterior vertebrae," 

 when one contemplates the extraordinary complexity of this apparatus and of the relations 

 of its numerous parts to each other, to the swim-bladder and to the otic region of the skull.' 



However, all this specialization has not prevented Sagemehl, Smith Woodward, 

 Goodrich, Tate Regan or any other author that I know of, from recognizing the probably 

 high antiquity of the Ostariophysi nor the primitive nature of many features of the jaws 

 and skull of the less specialized characins. 



Indeed if it were not for the possession of this apparatus the least specialized of the 

 Ostariophysi might well be classed among the Isospondyli. It is not impossible that the 

 family Lycopteridae from the Jurassic of China and Siberia, which are closely related to 

 the Leptolepidae, may stand in or near the line of ancestry to the carps, since the microscopic 

 characters of their scales, as studied by Cockerell (1925), approach the carp type. The 

 skeleton of the short-skulled Lycopterus sinensis, as restored by A. S. Woodward (1901, Pt. 

 IV, p. 3), would seem indeed to be a favorable starting-point for this order. According 

 to Cockerell, the anterior vertebrae of Lycoptera middendorffi. are not modified, so that the 

 Weberian apparatus had not yet been acquired in this very generalized forerunner of the 

 Cyprinidae. 



As might be expected, the otoliths of this order are widely different from those of 

 normal teleosts. Frost (1925, p. 553) states that in the suborder Cyprinoidea (of Regan, 

 including the characins, gymnotids and carps) the lagenar otolith, the asteriscus, is gen- 

 erally the most developed, often (except in the Cyprinidae) occupying lateral capsules or 

 open recesses in the cranial cavity above the floor supporting the brain. The utricular 

 otolith, the lapillus, is small in the Characiformes but well developed in the Gymnotiformes 



181 



