95. CH^TOSTOMUS. 243 



tends to the middle of the ventral. Caudal fin obliquely truncated. 

 Brownish : the upper surface of the head and the body with rather 

 irregidar yellowish spots, which are partly confluent into cross-bands. 

 Dorsal fin with three dark longitudinal bands. Lower parts whitish, 

 immaculate. (Kner.) 

 Barra do Rio Negro. 



5. Chaetostomus brachyurus. 

 Ancistrus brachyurus, Kner, Hypostom. p. 279. taf. 4. fig. 1. 

 D. 1/7. A. 6. L. lat. 23. 



Head depressed, without prominence on the occiput, as broad as 

 long, its length being one-third of the total (without caudal) ; eye 

 rather large, its diameter, being two-thirds of the width of the inter- 

 orbital space ; margin of the snout granulated, except in the middle 

 of its extremity, where it is naked. Interoperculum with about 

 twenty setiform spines, the longest of which are one-third as long 

 as the head. Scutes of the body with a serrated keel ; thorax and 

 belly naked; twelve scutes between anal and caudal. Dorsal fin 

 considerably longer than high, the length of its base being somewhat 

 less than its distance from the caudal. Pectoral spine longer than 

 the head, extending to the middle of the ventral. Caudal fin ob- 

 liquely truncated. Brown ; dorsal fin with three dark bands. {Kner.) 



Barra do Rio Negro. 



6. Chaetostomus itacua. 



Hypostomus itacua, Valenc. in d'Orb. Voy. Amir. Merid. Poiss.pl. 7. 

 fig. 1 ; Cuv. 8f Vol. xv. p. 505 *. 



D. 1/7. A. 5. 



Occiput with a slight prominence. Eye of moderate size, its dia- 

 meter being two-thirds of the width of the intei orbital space. Body 

 with light longitudinal bands ; fins with bluish eross-bands. (Val.) 



Rio de la Plata. 



7. Chaetostomus niveatus. 



Hypostomus niveatus, Casteln. Anim. Amir. Sud, Poiss. p. 43. pi. 21. 



fig. 3. 

 ? Hypostomus pictus, Casteln. I. c. p. 44. pi. 22. fig. 2 f. 



D. 1/7. A. 5. 



Dorsal fin considerably longer than high ; caudal obliquely trun- 



* This species is very imperfectly known, the figure given in d'Orbigny's 

 work being indifferently executed. Kner (Hypostom. p. 281) mentions a species 

 in the Berlin Museum, probably the same as that collected by Schomburgk in 

 British Guiana and determined by Midler and Troschel. From the notes made 

 by Kner from these specimens, it appears that they belong to a species different 

 from the itacua of the Kio de la Plata. 



t It will be better to bury this so-called new species of M. Castelnau among 

 the synonyms, than to carry it on as a doubtful form, as it is not probable that 

 it will be recognized and characterized at a future time 



r2 



