30 



ceiis it is said " L'opercule est etroit, et il y a quelques 

 ecailles sur sa surface, tandis que le sous-opercule et I'in- 

 teropercule en sont tout-a-fait depourvus :" [Hist, des 

 Poiss. iii. p. 52). If the specimens were uninjured, this 

 seems to be a sufficient distinctive character, but there 

 does not exist an equally strong reason for considering 

 salar as specifically distinct from Forster's trutta, though 

 the figures are not sufficiently detailed to prove their iden- 

 tity. I consider it, therefore, safer to retain the name of 

 salar, until the ichthyology of South Australia and New 

 Zealand has been more fully investigated, especially as the 

 authority of the Histoire des Poissons has consecrated a very 

 similar appellation to trutta, to a distinct species. Of G. 

 Forster's figures, the one numbered 211 in the volume, and 

 marked Scicena trutta (3, closely resembles salar in form, 

 and in the spots of the back forming transverse bars. 

 Number 210, which, like the preceding, is an unfinished 

 pencil sketch, was executed from a specimen taken in 

 Queen Charlotte's Sound on the 7lh of November, 1774, 

 and represents a more slender fish than our salar, with the 

 spots above the lateral line less uniformly round, and not 

 disposed in transverse rows. It shows also a longer soft 

 dorsal, with its last ray and that of the anal more ab- 

 ruptly produced ; the latter fin also is longer and more 

 even. Parkinson's figure, number 67, executed at Opoo- 

 ragee, in New Zealand, and 68, drawn in Queen Char- 

 lotte's Sound, seem to have been taken from fishes pre- 

 cisely similar in form to Forster's fig. 210, and therefore, 

 it may be concluded, of the same species, but differing in 

 the characters above-mentioned, from his 211. The names 

 inscribed on Parkinson's drawings are MuUoides sapidis- 

 simus, and Scicena inulloides, bestowed upon them by 

 Solander, whose notes on the species are referred to and 

 partly quoted in the Zoological Transactions (iii. p. 79). 

 Our figure of salar, which is very correct, will enable ich- 

 thyologists who may have an opportunity of examining 

 good collections of Australian fish, to clear up the obscurity 

 in which these species are still involved. 



We have elsewhere (p. 27) noticed the somewhat incon- 

 gruous assemblage of species in the Histoire des Poissons, 

 under the generic appellation of C'entropristes ; and the 

 authors of that work, as we have said above, mention the 

 Maenoid aspect of truttaceus, as an indicaUon of its being 

 the type of a distinct genus. Solander terms it, or a very simi- 

 lar species, MuUoides, and Forster says that trutta has many 

 claims to rank with the Mugiles, but taking the whole of 

 its characters into consideration, he was induced to place 

 it in the genus Scicena, which, in his time, was more com- 

 prehensive than it is now. The Centropristes georgianus, 

 which bears a close affinity to salar, has been chosen by 

 the Rev. Leonard Jenyns as the type of his genus Arripis, 

 so named because the scales of the body are destitute of 

 the usual fan-shaped furrows on their covered bases. In 

 salar the scales show distinct though not strong traces of 

 these furrows, as may be observed in the magnified figure 

 of a lateral scale (Plate XX. fig. 6), but we have, neverthe- 

 less, thought it right to place it in the same group with 

 f/eorgiamis, employing the word Arripis for the present, 

 merely as the name of a subgenus, from not being able to 

 determme satisfactorily the part of the system to which 

 the group ought to be referred. 



The form of Centropristes salar is described in the 

 Zoological Transactions as quoted above, and our figures 

 supply ample means of comparison with other species. 

 All our specimens have a space before and behind the eye 

 covered with a thick mucous deposit, resembling, in that 

 respect, certain states of the common mackarel, and the 

 disk of the preojjerculum is also veined in a manner not 

 very dissimilar to the same part in that fish. The speci- 

 mens vary from six inches to a foot in length. 



Hab. Bay of Islands, New Zealand (Sir James Ross) ; 

 Port Arthur, Van Diemen's Land (F. J. Lempriere, Esq.) ; 

 Queen Charlotte's Sound and Norfolk Island (J. R. Forster). 



Eleginds falklandicus. Richardson. 



Ch. Spec. El. preoperculo sub- an g id at o, pinna caudte 

 margins concavd. 



Radii:— Br. 6; D. 7|— 1|25; A. 1|2.3; C. 15|; P. 23; 



V. ijs. 



Plate XX., figs. 1, 2, natural size ; fig. 3 magnified. 



This Eleginus is the object of a considerable fishery at 

 the Falkland Islands, whence it is exported, after being 

 cured, to South America. In the forward position of the 

 ventrals and the pores on the jaw and head, this fish is 

 analogous to Notothenia, but its dentition is dissimilar, 

 and its lateral line continuous. The characters by which 

 the genus is discriminated from the other Scitenidce, 

 with which it is ranged in the Histoire des Poissons, are 

 there stated to be, — the entire preoperculum, small mouth, 

 long anal, very large pectorals, and jugular ventrals. Our 

 fish has these characters, and indeed auswers pretty closely 

 to the whole description of Eleginus niaclovinus, in the 

 work alluded to, but it differs from that and the other two 

 described species, in having fewer spines in the first dorsal. 

 I have had no opportunity of examining specimens of the 

 known species, but the plate of E. niaclovinus in the 

 Voyage de la Coquille (No. 17), fails entirely in giving the 

 generic aspect, and is manifestly inaccurate in the lateral 

 line, and in other particulars. The figure of the same 

 species, in the Histoire des Poissons (t. 115), gives the 

 general as])ect with more success, but differs from falk- 

 landicus in the distribution of the scales on the lower 

 limb of the preoperculum, in the form of the pectoral, which 

 does not coincide with the description in the text, in the 

 first three ra_vs of each of the vertical fins not being ap- 

 proximated to the other, and in the want of pores on the 

 head. We have no means of judging whether these dis- 

 crepancies be specific, or merely omissions arising from 

 inattention in the artist. 



Form compressed, fusiform, snout gibbous, head small. 

 Eye rather small. Posterior nasal opening placed nearer 

 to the end of the snout than to the eye. Anterior opening 

 very minute, and considerably before the hinder one. 

 Jaws moderately protractile. Maxillary small, and capa- 

 ble of being all retracted under the preorbitar, except the 

 lower corner. Mouth cleft about half way to the eye. 

 Teeth short, slender, rather obtuse, and erect, forming 

 narrow, not crowded villiform plates on the jaws. Tongue 

 and roof of the mouth smooth. Preorbitar and snout 



