some Loricariil FisJief. fj'i 



Pectoral spine cxtendiiif? to anterior ^ of ventral. Caudal 

 dcc|)ly eniar;;in:ite, the ine<liau rays not more than .', as 

 lonj; as the outermost, which are r-i the leri'jth of the fish. 

 Caudal peduuele 4-1^ as lon<; as deep, anteriorly with the 

 scutes of the lowest (fifth) series nearly flat and those of the 

 fourth scries strongly angulated. Numerous rather small 

 or oblonj; purplish spots on head, body, and fins, those on 

 the dorsal in 2 or 3 series on each interradial meuibraue. 

 Magdalcna and Cauca Jlivers. 



Professor Eigcnmann (' Science,' xxi. pp. 792-795) takes 

 exception to my name Plecostomiis Gunnani for the species 

 which he erroneously identified with P. lima, Liitken. lie 

 had given the name P. Inna atrojiinnis to a specimen, i)re- 

 sumably of this sijccies, with the fins uniformly dark brown, 

 and he says " Regan considers this variety distinct from lima 

 and names it Gunnani. Of course, if the variety atropinnis 

 is distinct from the present form it must go by the older 

 name atropinnis and not Garmani." This is hardly a correct 

 statt nicnt of the case. I reserve the name atropinnis for a 

 fish which I liave never seen, with the fins uniformly dark 

 brown *. This fish will probably prove to be specifically 

 identical with P. Gurnunii {P. lima (non Liitken), Eigeum.), 

 and if it be worthy of recognition as a variety, its name will 

 be P. Garmani, var. atropinnis ; if it should prove to be 

 specifically distinct from both P. Garmani and P. lima, its 

 name would be /'. atropinnis. 



The application of the generic name Ancistrus seems to 

 need some explanation. As defined and restricted by Gill 

 this genus includes species with an eight-rayed dorsal fin and 

 ■with the interoperculum armed with a tul't of rather long 

 and slender spines with hooked apices. This definition 

 appears to me applicaljlc only to A. brachyurus, A. scap/ii- 

 rliynchus, and A. medians of the species described by Kuer. 

 A. cirrhosus, afterwards selected by Bleeker as the type of 

 the genus, has rather short and stout interopercular spines, 

 curved towards the apices, which can scarcely be called 

 " hooked.'^ Dr. Gill's diagnosis was evidently ehiefiy based 

 on the specimens from Trinidad which he identified with 

 A. guacharote, Cuv. & Val., which Giinther named Clueto- 

 stomus trinitatis, and which have now been lost. In view of 



* I consulted ninro than one nutiiority as to the coiToctness of tliis 

 course, as I ciiil not wisli to dedicate a species to Dr. Uaiuian which 

 iui''ht be consideiLd a svnunvm. 



