IIYPOSTOMATINA. 221 



Sixth Subfamily. 



SILUIIID.E PllOTEROPODES. 



The rayed dorsal fin is always present and rather short; the 

 ventrals are inserted below (in all the genera, except Acestra), or 

 in front of, the dorsal (Acestra). The giU-membranes are confluent 

 with the skin of the isthmus, the giU-opening being reduced to a 

 short slit. Pectorals and ventrals horizontal. Vent before, or 

 not much behiud, the middle of the length of the body. 



Thirteenth Group. HYPOSTOMATINA. 

 Proteropodal Siluroids with the anterior and posterior nostrils 

 close together, and generally with a short flap between them. 

 The lower lip is reverted and much dilated, forming a broad flap 

 more or less deeply notched in the middle. 



The luiion of apparently such different genera as Arges, CalUchtliys, 

 Lorkaria, &c., is so opposed to the views of former ichthj'ologists, 

 that it will be necessary to add a few words in explanation. The 

 mailed body of the Loricarias and Hypostomes, reflected in some way 

 internally by the skeleton, is a verj' striking character : so much so, 

 that Kner is inclined to separate them from the Silm'oids altogether, 

 and Bleeker forms of them his first Siluroid family. However, that 

 the development of dermal sciitos is not a character of so much value, 

 is not only proved by Doras, where similar dermal productions cover 

 a part of the lateral line, and by the numerous mocUfications of the 

 dermal carapace on and behind the occiput, but especially bj' the 

 Iiichan genus Sisof, which has a series of dorsal plates, the sides 

 being almost naked. With regard to the skeleton, the compression 

 of the caudal vertebra) and the dilatation of their neural portion are 

 not only found in Loricarhi and Hypostomus, but also in other slender- 

 tailed Siluroids, as Sisor, Bunocephalus, and Aspredo, genera which we 

 associate with the former, but which are excluded from the Lorica- 

 roids of Kner and Bleeker. 



Eveiy doubt that Loricaria and the genera allied to it are true 

 SUm'oids, and nothing but Siluroids, must disappear when they are 

 compared with those ^\^th which they have been associated by 

 myself. There we find the gQl-opcning reduced to a foramen or 

 short slit, in consequence of the gill-membranes being confluent 

 with the skin of the broad isthmus ; the nasid openings placed close 

 togetlier in a small groove ; the ventrals inserted far forwards, 

 below, or even in front of, the short dorsal ; the pectorals and ven- 



