SOUTH AMERICAN NEMATOGNATHI. 401 



appeared emarginate. Dr. Steindachner has identified 

 horridus with this species. 



On page 286 Prof. Cope states that scojndarius differs 

 from his biseriatus in having the dorsal radii, ii, 12.1. 

 The plate, however, shows the dorsal radii as i, 7.1, and 

 a comparison of Cope's and Kner's plates will convince 

 any one of the identity of the two species. 



Dr. Steindachner separates his tenuicauda on account 

 of the entire granulation of the snout. We have exam- 

 ined a number of specimens which are undoubtedly 

 identical with einarginatus which have the tip of the 

 snout granular, and the type of emarginatits will most 

 probably be found to have a granular snout. While 

 this is usually a convenient character in separating 

 species, it is not constant in einargi7iatus and plecostomus^ 



We would include under einarginatus all those indi- 

 viduals in which the tail is broad, flat below, margined 

 on the sides by the posthumeral keel, and which have 

 the peculiar arrangement of plates already described. 

 Of how little value other characters are may be seen 

 from the following comparisons: 



Two specimens of einarginatus, each .215 m. long, one 

 a 3 from Fonteboa, the other a ? from Hyavary, differ 

 considerably. 



MALE SPECIMEN. 



Occipital ridge or keel prominent. 



Nuchal plates bicarinate. 

 All the lateral plates keeled. 



Tail more slender than in $ . 



Snout pointed, entirely granular. 



Pectoral spine reaching beyond 

 middle of base of ventrals. 



Much darker colored than the fe- 

 male. 



26 



FEMALE SPECIMEN. 



Occipital keel or ridge almost obso- 

 lete. 



Nuchal plates not bicarinate. 



Lateral plates on the abdominal 

 portion without a trace of a keel, 

 some of those on caudal portion 

 slightly keeled. 



Snout more rounded, naked at tip. 



Pectoral spine not reaching base 

 of ventrals; all the fins lower than 

 in the 3 . Pectoral spine and out- 

 er caudal rays bristly. 



