[3] REVIEW OF THE SCIiENID.E. 345 



Scicena, is ho better, as very many of the species have this spine neither 

 large nor small, and could as well be placed in the one group as the 

 other. Bleeker divides this group into Pseudo scicena, species with the 

 mouth oblique and the jaws subequal, the lower jaw with the teeth of 

 the inner row enlarged, and Johnius with the mouth horizontal and the 

 lower jaw included, the teeth of the lower jaw being in villiform bands. 

 This arrangement is better than the other only in theory. The charac- 

 ters chosen are of more value as indicating relationship, but they cannot 

 be applied in practice, as there are intermediate gradations of all sorts. 

 The type of Pseudosciccna {Scicena aquila) is in fact much more nearly 

 related to the type of Johnius than to most of the species associated 

 with it in Pseudosciama. 



As we proceed along the series of Sciceninw from Larimus towards 

 Menticirrhus, the following changes are notable : In the Larimus type the 

 pores on the snout are small and few, and there are no distinct slits or 

 lobes on the snout above the upper jaw; in the other type the pores be- 

 come large and conspicuous, 4 to 6 in number, and the thickened snout 

 above the upper jaw has two slits on each side, bounding two dermal lobes. 

 The mouth becomes smaller, narrower, more horizontal as we proceed 

 towards Menticirrhus, the lower jaw shorter, and the bands of teeth in 

 both jaws more and more broad, those in the lower more decidedly vil- 

 liform ; the pores on the chin become larger and more numerous, the 

 number rising from 2 to 5 ; the lower pharyngeals become larger, and 

 their teeth larger and less acute ; the preorbital becomes wider and 

 more gibbous, the gill-rakers shorter, fewer, and more like tubercles ; 

 the anal fin is placed farther forward, and the spines of the fins gen- 

 erally are less slender ; the scales, as a rule, become rougher, and the 

 rows of scales less regular in their direction. The flesh, as a rule, be- 

 comes firmer, coarser, less agreeable in flavor, and of less value as food, 

 but this, like some of the other characters mentioned above, is subject 

 to much variation. 



It may be noted that in some Scicenidce the middle rays of the caudal 

 are more produced in young specimens. In some also the serrations 

 on the preopercle become weaker or even obsolete with age. 



The two subfamilies recognized by us may be thus distinguished : 



ANALYSIS OF THE SUBFAMILIES OF SCI&NIDm. 



a. Vertebras typically 14 + 10, the number in the abdominal region always greater 

 than that of the caudal; lower jaw prominent; teeth not villiform; edge of pre- 

 opercle entire ; second anal spine weak and adnate to the first ray ; the first spine 

 minute and often obsolete Otolithin^e , I. 



aa. Vertebrae typically 10 -f- 14, the number in the caudal region always greater than 

 that in the abdominal ; second anal spine usually well developed and usually joined 

 to the first soft ray by a distinct membrane Sci^nin^e, II. 



