OBSERVATIONS ON FISH SCALES. 
165 
similar, all purely cycloid, with the acanthopterygian type of basal radii. They probably represent a 
cycloid derivative from the ctenoid series, but they certainly can not be derived from the gerrid type I 
have examined. The nucleus is central or slightly above, and the basal radii are spread out fan-wise. 
The basal radii are about 15-19 in Phanerodon, 11-13 in Damalichthys , 9-10 in Zalembius, 7-10 in Amphis- 
ticus. The lateral circuli are angulate in Amphisticus and obtusely bent in Zalembius; in the large 
(8 mm. diam.) scales of Phanerodon the bending of the lateral circuli is evident near the middle, but 
becomes evanescent toward the periphery. The apical and basal cir¬ 
culi in Phanerodon are very fine and dense. 
On comparing these scales with those of the Pceciliids, which ap- 
peaf to be primitively cycloid, it is not at first evident that there is 
any characteristic difference. The Pceciliids do not show the bending 
or angulation of the lateral circuli, and their system of basal radii is 
less regular and fan-like. The general facies is distinctly different, 
yet it is difficult to point out good diagnostic characters capable of 
being expressed in words. The atherinid scale, as one would expect, 
ranks rather with the poeciliid than the embiotocid. 
Labrid and scarid scales, although cycloid, are entirely different from all these. Jordan places the 
Embiotocids in a separate suborder, Holconoti, remarking that “it contains fishes percoid in appear¬ 
ance, with much in common with the Gerrids and Sparidae, but with certain striking characteristics not 
possessed by any perch or bass. * * * The lower pharyngeals are solidly united, as in the Labrids, 
a group which these fishes resemble in scarcely any other respects. ’ ’ 
It seems just possible that the Embiotocidae, although not close to the Labridae, may agree with them 
in having had (as I certainly must suppose of Labridae) no ctenoid ancestors. In this case they would 
apparently be distant relatives of the scombroid and atherinoid series. We have, however, in the 
Cichlidae a scale development similar to that of the Embiotocidae, and certainly from ctenoid ancestors. 
(Compare the scale of Tilapia, figured in Smithsonian Miscellaneous Collections, volume 56, no. 1.) 
Suborder Chromides 
Here we return to the normal percoid type of scale. 
POMACENTRIDiE. Coral-fishes. 
I have examined the cockeye pilot, Eupomacentrus leucostictus (Muller & Tro- 
chel), from Key West, Fla. (W. C. Kendall), and Abudefduf saxatilis (Linnaeus), from 
Socorro Island (Albatross). The scales look like those of the Sparidae. They are 
broader than long, those of Eupomacentrus being extremely broad (long. 3, lat. nearly 
5 mm.). The basal radii are about 9 to n. In both genera the lateral circuli are 
slightly oblique. This is a case in which the fishes have developed very marked 
structural characters, without any material evolution in the scales. 
CICHLIILE. Cichlids. (PI. xxxix, fig. 39, 40, 41-) 
The cichlid scales much resemble those of the pomacentrids, but the lateral cir¬ 
culi are not oblique, and the submarginal apical elements, instead of being truncate, 
are obtusely pointed and spinelike, herein approaching the condition of the Ana- 
bantids. It thus becomes possible to get an idea of the evolution of the complicated and specialized 
ctenoid patch of the percoids, the stages being represented more or less accurately by the Macruridae, 
Anabantids, Cichlids, and Pomacentrids, although, of course, this arrangement does not correctly 
represent the evolution of these particular families. It is also worthy of note that Tilapia mlotica 
(figured in Smith. Misc. Coll., 56, pi. h fig- 3) has exactly the anabantid type of circuli, with widely 
spaced coarse ones at the sides of the apical area. The middle of the apical area in T. mlotica is covered 
Fig. 37. —Abudefduf 
saxati lis (Poma- 
centridae). Cte¬ 
noid structures. 
Bureau of Fish¬ 
eries. 
Fig. 36. —Eupomacentrus leucostictus 
(Pomacentridae). Ctenoid struc¬ 
tures. Bureau of Fisheries. 
