observations on fish scales. 
125 
The character given in Jordan and Evermann’s key for all the clupeines except Brevoortia, “scales 
with their posterior [apical] margins entire and rounded” is not accurate, although Brevoortia scales are 
readily separated from the others by their long weak teeth. 
PHRACTOKFMID#. 
Phractolcemu 9 is a living African fresh-water genus; Jordan has the family printed by error in italics, 
as extinct. The scales of the only known species are described in the Proceedings of the Biological 
Society of Washington, volume xxiii, page 112, and figured in Biological Bulletin, volume xx, plate 
hi, figure 17a. They strongly suggest certain of the Characinidae. 
OSTEOGLOSSIDiE. Bony-tongued fishes. 
There are at least two subfamilies, Osteoglossinae ( Osteoglossum and Scleropages ) and Heterotinae 
(Heterotis ). The scales of this group, remarkable for their regular network and beaded circuli (primi¬ 
tive characters recalling the Dipneusti), have been discussed in the Proceedings of the Biological Society 
of Washington, volume xxiii, page hi, and volume xxiv, page 39; Science, May 26, 1911, page 831; 
and Heterotis is figured in Biological Bulletin, volume xx, plate hi, figure 20. Mr. W. M. Mann has 
kindly sent me scales of the Amazon bone-tongue, Arapaima gigas (Cuvier), which Eigenmann makes 
the type of a distinct family, Arapaimidce. They are about 78 mm. long and 55 broad, superficially 
like those of Neoceratodus. The exposed part has a very coarse labyrinthoid rugosity. The concealed 
part has elongated reticulations and beaded circuli; the reticulation very irregular and not nearly so 
well developed as in Osteoglossum. The scale is really very near to that of the Nile bone-tongue, 
Heterotis. 
PANTODONTIDAJ. Pantodonts. 
Living African fishes, only one species known. The scales are described in the Proceedings of the 
Biological Society of Washington, volume xxm, page 112, and figured in Biological Bulletin, volume 
xx, plate m, figure 18. The circuli are strongly beaded. 
N0T0PTERIDA5. Finbacks. 
Fresh-water fishes of the Ethiopian and oriental regions, considered by Boulenger to be nearest to 
the Hiodontidae. The scales of the African Notopterus afer Gunther are described in the Proceedings 
of the Biological Society of Washington, volume xxiii, page 112. In this place I have remarked on the 
resemblance of the scales to those of Gadus. This is certainly true as regards shape and general appear¬ 
ance, but on close comparison it is seen to be wholly superficial. In the gadines the numerous fine 
radii extend in all directions from the nucleus; in Notopterus, except for a series of striae just below the 
apical margin, they are wholly basal, and only the central ones reach the nuclear region. In Notopterus 
these radii are only about half as dense (close) as in Gadus, but on the other hand the circuli are much 
denser. The circuli of Notopterus are not at all beaded. 
SALMONIDAJ. Salmon, trout, etc. 
Very good figures of Salmo scales have been published by Mr. J. Arthur Hutton in his pamphlet on 
“Salmon Scale Examination and its Practical Utility” (London, 1910). He shows that the life of the 
fish affects the growth of the scales, so that by examining the scales it is possible to determine the 
approximate age, the time spent in the river and in the sea, and other important facts. The salmonoid 
scales before me are the Atlantic Salmon, Salmo salar Linnaeus, Woods Hole, Mass.; the golden trout, 
S', roosevelti Evermann, cotype, Mount Whitney, Cal.; Coulter’s whitefish, Coregonus coulterii Eigen¬ 
mann & Eigenmann, cotype, Kicking Horse River, Field, British Columbia; the brook trout, Salvelinus 
fontinalis (Mitchill), Woods Hole, Mass.; the blueback or red salmon, Oncorhynchus nerka (Walbaum), 
Wallowa Lake, Oreg. (J. J. Stanley). I have figured the scale of the blackfin, Leucichthys nigripinnis 
(Argyrosomus nigripinnis Gill), in Biological Bulletin, volume xx, (1911), plate 1, figure 2. The scales 
of Salmo are more or less oval, with well developed concentric circuli, which variously fail in the apical 
region; the radii are absent. Scales of Salvelinus and Oncorhynchus are essentially the same; with 
