210 SILUKID^. 



serrated. The upper and lower surfaces of the tail are shielded 

 behind the adipose and anal fins. Humeral process striated, sword- 

 shaped, broader than, and extending to the hinder fourth of, the 

 pectoral spine. Eye small, covered by a membrane, as distant from 

 the end of the snout as from the gill-opening. The barbels are free, . 

 not united at the base, provided with very small tentacles ; those of 

 the maxillaries do not extend to the gill-opening. Teeth very small, 

 present in both jaws. Dorsal and caudal fins with dark spots. (Kner.) 

 Eio de la Plata. 



83. SYNODONTIS. 



Synodontis, Cm: ^ Val. xv. p. 244. 



Bracliysynodontis, Synodontis, Pseudosynodontis, Ilemisjaiodontis, 

 Leiosynodontis, Bleek. Nedeti. Tydschr. Dierk. I860, p. 52. 



Adipose fin of moderate length, or rather long ; dorsal with a strong 

 spine and ■with seven soft rays ; anal rather short. Teeth in the 

 lower jaw moveable, very thin at the base, and with a slightly dilated, 

 pointed, brown apex ; palate edentulous. Mouth small, mandibles 

 short. Barbels sis, more or less fringed wth a membrane or with 

 filaments. Eyes of moderate size. Neck with broad dermal bones. 

 Tlie gill-openings are of moderate width. Yentrals with seven rays, 

 inserted behind the dorsal. 



Tropical Africa. 



Knowing most of the species of Si/nodonfis from autopsy, I cajinot 

 adopt the views of Dr. Bleeker, who has split this natural generic 

 group into five divisions : — 



a. Synodontis has been restricted by him to a species with long, 

 slender intermaxillary teeth and with ciliated maxillary barbels (S. 

 macrodon). The discovery of a new species {S. sorex) already shows 

 that the characters used for the foundation of this genus are unser- 

 viceable : either it must lose one of the characters on which it is 

 based, or S. sorex would become the type of a sixth genus. 



h. Pseudosynodontis is created for S. serratus, a species most 

 closely allied to S. schid, which, however, is referred to Hemisyno- 

 dontis. Although I am fully convinced of the distinctness of the 

 two species, I confess that, seeing the great variations in the develop- 

 ment of the spine of S. sclial, I at first entertained some doubts about 

 it, nor did I feel satisfied on this point until I had ascertained by 

 dissection that the one was not the male sex of the other. If these 

 be considered tj'pes of two genera, then we may similarly separate 

 the short- and long-snouted Eels of our rivers. 



c. Hemi synodontis is created for S. memhranaceiis, S. schal, and S. 

 nigritus. According to Dr. Bleeker's \'iews, this is an assemblage 

 of most heterogeneous fishes, — the S. memhranaceus differing from 

 the others bj' a great extent of the gill-opening, and by a structure 

 of the maxOIary barbels, which is at least of as much importance as 

 the " cirri fimhriati " of Synodontis, Blkr. 



d. Leiosynodontis is founded on S. macvdosus, the younc/ of Hemi- 

 svnodontis scTud. 



