8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 56 



2. Broad scale with very many apical radii; nuclear area granular; scale 



about 7 mm. long and 10 broad, the largest of the Citharinine 



scales Citharinus gibbosus Boulenger. Niamkolo, 



Lake Tanganyika (Cunnington) . 



Rounder scales, with fewer apical radii; nuclear area with circuli to 



middle 3 



3. Scale larger, nearly 6 mm. long, about 7 broad ; character of sculpture 



entirely as in C. gibbosus Citharinus macrolepis Boulenger. 



Boma, Lower Congo {Delhez). 

 Scale smaller, about 5 mm. broad; otherwise practically the same. . 



. . Citharinus latus Muller and Troschel. 

 Beni Souef (Loat). 



4. Nuclear area with circuli to middle ; apical area with very large, coarse 



circuli, but not quite abruptly separated from lateral; scales small, 

 a little over 3 mm. wide. .. .Citharinus citharus Geoffroy (geoffroyi 

 Cuvier). White Nile (lake near) (Loat). 

 Nuclear area broadly granular ; apical circuli coarse ; scale 4.33 mm. broad. . 



. . Citharinus congicus Boulenger. 

 Bolobo, Congo (Grenfell). 



The above key is not wholly satisfactory, because the scales of Citharinus 

 are really of a very uniform type, without very obvious specific characters. 

 Two tribes are indicated : Citharidiini for Citharidium, and Citharinini for 

 Citharinus. 



Boulenger places two American genera in this subfamily, but Gill separates 

 them as Curimatinse. I have not seen their scales. 



Hydrocyon forskalii runs in the above table near Citharinus congicus, but 

 the resemblance is merely superficial. 



ICHTHYOBORIN^ 



Carnivorous forms with ctenoid scales. There is nothing to distinguish 

 their scales from those of the Distichodontinse. The apical teeth are numerous 

 and close together, having little in common with those of Citharidium. The 

 species examined are : 



Phago boulengeri Schilthuis. Monsembe, Upper Congo (/. H. Weeks). 



The scales are hard, bone-like plates joined together, but the apical 

 densely-set spine-like teeth are quite as in the other forms. 

 Eugnathichthys cetvcldii Boulenger. Monsembe, Upper Congo (J.H. Weeks). 

 Small subquadrate scales, with the nucleus apical, just below the row of 

 teeth ; no radii, but the basal region is broadly plicate. 

 Ichthyoborus besse Joannis. Fashoda {Loat). 



Scales practically as in the last. 

 Neoborus ornatus Boulenger. ]\Ionsembe, Upper Congo (/. H. Weeks). 



Scales also as in Eugnathichthys , except that the apical region is coarsely 

 rugose. 



Paraphago and Mesoborus have not been examined. 



