702 



THE GENUS GNATHANACANTHUS GILL. 



to the Cataphracti, and especially the Scorpwnoidei in the following 

 terms: 



" La chaine sousorbitaire y est eu effet complete et s'articule avec le 

 preopercule, mais les os sont rudimentaires en ce sens, qu'ils formeot 

 des plaques tres-minces, dont le posterieure, de forme oblongue et de 

 presque la longeur de I'orbite, se retr^cit eu arriere pour s'y articuler 

 avec le preopercule vers le milieu de la hauteur de son bord post^- 

 rieur." 



In 1878, Count Castelnau noticed a fish which he referred to a new 

 genus called Beridia or Baridia {B. Jiava)* and which he referred to 

 the family Triglidw, with the following explanation : 



"This new genus belongs to the Triglidw, and its spinous dorsal 

 being rather less developed than the soft, ought probably" to be placed 

 in the group Cottitia, but the general form is very different from all the 

 other fishes of Triglida^ and is more like some sorts of Gohiidcey 



In 1879, the present author, in a brief summary of the progress of 

 Vertebrate Zoology in 1878, referred to Beridia in the following terms : 



" Quite a large number of new genera of fishes have been proposed, 

 but several of them are unquestionably the result of imperfect knowl- 

 edge or erroneous ideas, and among such may be mentioned those 

 named by Count Castelnau (1), Brishania and (2) Baridia or Beridia. 

 The former was proposed for a fish occurring in the Brisbane Eiver, 

 and is undoubtedly identical with Megalops, while the latter is the 

 same as Gnathanacanthus, long before described by Bleeker." t 



In 1883, Mr. Eobert M. Johnson, in an ably compiled catalogue of 

 the fishes of Tasmania, enumerates Holoxenus cutaneus as one of the 

 species of Triglid(e, and added the following comments: 



" I have not seen the above, but I have good reason for supposing 

 that the fish, not otherwise mentioned, known as the Velvet Jish, is 

 probably the same, although the spinous characters are not in agree- 

 ment with those of II. cutaneus.''^ 



A description is then given of the Velvet fish, and it is added, 

 '< Should it prove to be a distinct species I propose for it the name 

 Eolexenu.s Gnntheri.'''' | 



n. 



The nominal genera and species thus introduced are undoubtedly 

 congeneric. Whether they are based on the same species is not so clear. 

 The differences of the radial formula are considerable, viz : 



D. 



A. 



Gnathanacanthus Goetzi 



Holoxenxtg cutanetit 



Beridia flava 



Holoxenus Qiintheri 



71 51 n 



7 I 3 I 10 

 §8 I 3 I 10 



8 I 5 I 10 



3 9 



[3J.9 

 2.8 

 3.9 



12 



12 



12-13 



10 



*0n the plate (2) the name is printed Baridia flava. 



t Ann. Eec. Sc. and lud. 1878, p. 458, 1879. 



: Papers Royal Soc. Tasm., 1882, pp. 114, 115, 1883. 



§ In specific diagnosis, 7 spines ; in generic, 8 ; in figure, 8. 



