SALMONID^. 137 



of the upper edge of the orbit, diameter 3 to 3^ in the length of the head, 

 I diameter from the end of the snout, and also apart. The shape of the fish is 

 peculiar, the back, sides, and abdominal surfaces being flattened so as to give it a 

 general tetragonal appearance, while keeled ridges divide the various surfaces one 

 from the other. These four surfaces are further subdivided by other parallel ridges, 

 one of which is a short distance internal to the superior orbito-caudal ridge, and 

 another superior to the pectoral ridge. Irrespective of these, a short ridge passes 

 from the lower edge of the base of the pectoral fin to the anal fin. Upper jaw 

 slightly the longer : the posterior extremity of the maxilla scarcely reaches more 

 than two-thirds of the distance to beneath the front edge of the eye. Snout 

 depressed and somewhat pointed : upper surface of the head flat, its sides 

 compressed. The suborbital ring of bones, the preopercle, opercle, and upper 

 portion of the subopercle with a rather thick adipose covering. Teeth — none 

 in the jaws : a curved row of small ones across the vomer near its head, and 

 continued on to the anterior portions of the palatines, which are contiguous : 

 a single row of about eight large and somewhat recurved ones surround the 

 inner edge of the front portion of the tongue. GUI-openings deeply cleft. 

 Gill-rakers rather widely separated, thick, and equalling in length about 

 one-fourth of the diameter of the orbit, i^ms— the first dorsal anteriorly as 

 high as the body below it, its posterior rays rather more than two-fifths the 

 height of its anterior ones : adipose dorsal fin situated on a line above the last 

 anal ray. Pectorals, if turned forwards, reach to the middle of the eye. Ventrals 

 inserted below the last dorsal ray and midway between the end of the snout and 

 the base of the caudal fin. Anal highest anteriorly where the rays equal the length 

 of the base of the fin. Caudal forked. Scales — large, thin, and higher than 

 broad, those along the back adherent, those on the sides more deciduous. Minute 

 ossicles, assuming a stellate or spinate form, are present on the scales of the back 

 and on some of those along the under surface of the abdomen. Those on the 

 sides with roughened or ragged outer margins : the row beneath the lateral-line 

 the largest. Numerous concentric lines upon the surface of the scales. Lateral- 

 line — on a row of scales somewhat smaller than those above or below, it is well 

 marked, and passes to the centre of the base of the caudal fin. Air-bladder — 

 smaller than in the trout, from which it also differs in the thickness of its fibres 

 and silvery walls, the latter charged with a brilliant substance that separates with 

 washing, at first in the form of spangles. This is next precipitated by ammonia 

 into a silvery deposit, which is preserved for the manufacture of artificial 

 pearls. Colours — in an example examined a few days after death, but which had 

 been preserved in spirit, it was of a light olive along the back, becoming silvery 

 white along the middle of the sides, while inferiorly and on the head there was 

 a tinge of gold glossed with purple. A black spot on the upper edge of the orbit, 

 and a smaller one on the snout. A dark longitudinal mark along either caudal 

 lobe near to its outer edge. Edward, who received a fresh one, observes that 

 the colours agree with the number of divisions on the fish. That along the back 

 deep amber, the succeeding one greenish-blue, the third or division below the 

 lateral-liue, of a pure silvery white with bright metallic lustre, the others similar 

 to those above but fainter. Belly grayish-white with a tinge of greenish-blue and 

 purple. The top of the head nearly the same as the back, its sides silvery. Eye 

 with a bright ^silvery iris and blusOi-black pupil. It was so transparent that on 

 holding it up to the light the vertebras could be distinguished. 



The fish is figured and described by Rondelet, and this was reproduced by 

 Gesner, while Willughby observes " Fisciculus Romre Argentina dictus, Sphyrcena 

 parva, sive Sphyrcena secunda species, Rondeletio, Gesner, 1061." This clearly 

 shows that the species of these three authors were identical with the Roman fish, 

 the air-bladder of which was employed to furnish materials for the manufac- 

 ture of artificial pearls. Eay copies, almost verbatim, from the above authors. 

 Artedi's Argentina is identical with that of Linnreus, but not with Gronovius's 

 Sphyrcena. Gmelin in his edition of Linnteus possibly mixed up, under the 

 designation of Argentina sphyrcena, more than one species of fisb ; still it is clear 

 that the older authors referred to the deep sea fish, from which materials for false 

 pearls were procured, as stated by Risso. 



