bulletin of the bureau of FISHERIES. 
168 
having perhaps coalesced. The basal radii are weak, about 3 to 7, and the excessively dense lateral 
circuli are not at all oblique. 
The fishes represent two subfamilies, Chaetodontinae and Pomacanthinae. 
• Suborder SclERODERMi. Hardskin fishes. 
BAUSTIDAE. Trigger-fishes. 
The leather jacket, Balistes carolinensis Gmelin, has peculiar thickened, extremely broad scales, 
about 3X mm. long and 7 broad, the apical half covered with large tubercles, which appear to be homol¬ 
ogous with the spines of the Macruridae. These tubercles or spines are curiously suggestive of those on 
Fig. 38. —Chatodon ulie- 
tensis (Chaetodonti- 
dae). Ctenoid struc- 
tures. Bureau of 
Fisheries. 
Fig. 39. —Pomacanthus arcuatus 
(Chaetodontidae). Ctenoid 
structures. Bureau of Fish¬ 
eries. 
Stanford University. 
the scutes of Acipenser. Those along the basad margin of the tuberculate area are arranged in a regular 
row, which is angled in the middle, the very broad angle looking apicad. A quite different looking 
Balistes scale, with six large spines, is figured in the Cambridge Natural History, Fishes, page 191. 
MONACANTHIDAE. Filefishes. 
A scale of Monacanthus , of the same general type as that of Balistes, is figured in the Cambridge 
Natural History, on the page just cited. The large spines are arranged much as in the macrurids (e. g., 
Macrurus carminifer Garman). 
I have material of the orange filefish, Ceratacanthus schcepfii (Walbaum), from near Woods Hole, Mass., 
kindly given to me by Dr. E. Linton. The skin contains many minute spots of black pigment, and in 
places a fine dark blue pigment, which does not turn red in acids. The scales are shaped nearly as in Mon¬ 
acanthus and the exposed area of each has numerous large spine-like structures, one of which is much 
larger than the others. The concealed part of the scale is striate, with the appearance under the micro¬ 
scope of tree trunks closely packed, branching more or less distally. 
