OBSERVATIONS ON FISH SCALES. 
147 
lines, so as to seem segmented. In 5 '. borealis DeKay, from Woods Hole, the colorless scales have a 
diameter of hardly iK mm., and the fine radii, instead of continuously interrupting the circuli, are 
represented by series of minute round holes, which, however, coalesce in some places, producing a 
condition like that of the great barracuda, 5 . picuda. In the northern barracuda, 5 . borealis, the apical 
sculptureless area is very small or wanting. The sphyranid scale is very suggestive of that of Gadtis 
and allied genera. 
Ac"' A 'A" 
A;^.A' A 
A '^A 
A 
A A 
Fig. 9. — Mug'll (MugilidEe) . Cte- 
noid area. Bureau of Fisheries. 
Fig. 10. — Spkyrtena borealis (Sphyraenid^). Sculpture. 
Bureau of Fisheries. The transverse strands are circuli. 
Jordan & Evermann state that the families Atherinidae, Mugilidse, and Sphyraenidse are closely 
related (Bulletin 47, U. S. National Museum, pt. i, p. 788). They were associated together in the 
order Percesoces by Cope. Boulenger includes in Percesoces several other families, as Anabantidse, 
Stromateidse, Polynemidse, Scombresocidse, etc., stating that the group is perhaps only an artificial 
one, but “a gradual passage may be traced connecting the most aberrant types.” 
The scales would certainly suggest that the three families described above are not very closely 
related. 
Suborder Rhegnopteri. 
POLYNEMID.®. Threadfins. 
Jordan states that the Polynemidae are allied to the Mugilidse, but differ from them and from all 
other fishes in the structure of the pectoral fin and its basal bones. In Boulenger ’s arrangement they 
go in the Percesoces, following the Mugilidse. In Polydactylus octonemus Girard the scale is quite 
Fig. h. — Polydactylus {VoXyaerm- 
d$). Apical teeth. Bureau of 
Fisheries. 
Fig. 12. — Per copsis guttatus iPercopsidx). Marginal teeth. Bureau 
of Fisheries. 
typically Acanthopterygian, with ctenoid apical area, nucleus apicad of middle, and well-developed 
basal radii. The scale is nearly as in Mugil, having the same basal emargination, but differing in the 
spreading basal radii. The minute elements of the apical area are not as in Mugil, the submarginal 
ones being truncate instead of pointed. 
Although there are differences, the scales would suggest that the Polynemidse are actually nearer 
to the Mugilidse than the latter to the Atherinidse or especially the Sphyrsenida. 
