122 
bulletin of the bureau of fisheries. 
Order HALECOMORPHI. 
AMIIDAi. Bowlins. 
I still write Amia and Amiidae for the bowfin, being unconvinced that the rules of nomenclature 
require the change which has been proposed. The scale of Amia has been discussed in several of my 
earlier papers (e. g., Smithsonian Miscellaneous Collections, vol. 56, no. 1, p. 1). It has simple longi¬ 
tudinal fibrillae, and no radial network. In its general structure it is not far from Ccelacanthus, or indeed 
very remote from Neoceratodus and Sagenodus. No comparison is possible between the scales of Amia 
and Lepisosteus, and it is Amia alone which points toward the Teleostei. 
Superorder TELEOSTEI. The bony fishes. 
Order ISOPONDYLI. The isospondylous fishes. 
I would change Jordan’s arrangement of families to the extent of placing the Albulidae first among 
the living groups, as undoubtedly the most primitive. It alone has the strictly longitudinal basal 
Fig. i. —Amia calva. 
Fig. 2. —Albula vulpes 
(Albulidae). Bureau 
of Fisheries. 
Fig. 3. —Elops saurus (Elopidae). 
Bureau of Fisheries. 
fibrillae, which are tuberculate as in the Dipneusti. It alone, also, has two transverse series of valves 
to the bulbus arteriosus instead of one, another primitive character. 
ALBULIDAE. Bonefishes. 
The scale of Albula is shown in text figure 2; that of Dixonina does not differ. In outline and 
general form these scales resemble those of Synodus and Lucius or Esox (pi. xxxv, fig. 19, 20), but 
these latter have transverse basal circuli. 
ELOPIDAJ. Ten-pounders. 
For a discussion of the scales of Elops see Smithsonian Miscellaneous Collections, volume 56, 
no. 3, page 3. Their structure shows nothing especially primitive; the basal circuli are very finely 
beaded, the lateral ones simple. The apical circuli, above the nucleus, break up into transverse 
labyrinthiform markings, the sculpture much coarser than that of the normal circuli. 
MEGALOPIDvE. Tarpons. 
Jordan makes a separate family for the tarpons, although on an earlier page of the same book he 
follows the more usual custom of referring them to the Elopidae. I am indebted to Mr. V. N. Edwards 
for a scale of Tarpon atlanticus (Cuvier & Valenciennes), from Woods Hole, Mass. It is of course very 
large (62 mm. long, and almost as broad), and is formed much as in Elops, differing, however, by having 
only three basal radii (12 to 20 in Elops), the basal margin broadly scalloped much as in Synodus. The 
