152 
BULLETIN OF THE BUREAU OF FISHERIES. 
Subfamily LEPOMIN.E. 
Tribe Ambloplitini. (PI. xxxvii, fig. 29, 30.) 
Strictly cycloid, but the apical circuli widely spaced, and more or less strongly undulate or zigzag, as 
in Gerridae Acantharchus pomotis (Baird), Tarboro, N. C. 
Sharply ctenoid when young, but margin cycloid in adults Ambloplites rupestris (Rafinesque) 
With a small variously developed ctenoid area Chcenobryttus gulosus (Cuvier & Valenciennes) 
Very strongly and coarsely ctenoid Arckoplites interruptus (Girard) 
Thus Centrarchus stands at the base of the Centrarchinae, and Arckoplites at the base of the Amblo- 
plitini. 
Tribe Lepomini. (PI. xxxvii, fig. 28.) 
For the pharyngeal characters, see Bean & Weed, Proceedings United States National Museum, 
volume 40, pages 369-376. On the scale characters, Enneacanthus stands at the base of this series. 
Scales strongly ctenoid Emieacanthus gloriosus (Holbrook) 
Scales moderately ctenoid, not so broad as those of Enneacanthus. 
Mesogonistius chcetodon (Baird) Trenton, N. J. 
Scales feebly ctenoid, rarely practically cycloid, with no generic difference from the last, except that 
they are usually less ctenoid and have fewer basal radii Lepomis 
It seems impossible to make any key for the scales of the species of Lepomis; L. gibbosus, cyanellus, 
pu 7 ictatus, auritus and some euryorus have the apical circuli more widely spaced than in the others. 
The species examined are L. cyanellus, punctatus, symmetricus , auritus, megalotis, pallidus, albus, keros, 
kolbrooki, euryorus, gibbosus. The scale of L. megalotis (Lake Maxinkuckee, Ind.) is quite broad, 
with up to 12 basal radii, a larger number than is usual in the genus. 
Subfamily MICRO PTERINAJ. Black basses. 
The genus Micropterus has been discussed above. The scales are the largest foimd in the Cen- 
trarchidas, but Ambloplites runs them close. Ambloplites differs from Micropterus in having the adult 
scales conspicuously broader than long, with denser apical circuli. It is a question whether M. salmoides 
should not be called Aplites salmoides, or Aplites Rafinesque at least used in a subgeneric sense. 
The following arrangement of the centrarchid genera is suggested: Centrarchus, Pomoxis; Aplites, 
Micropterus; Arckoplites, Chcenobryttus, Ambloplites, Acantharchus; Enneacanthus, Mesogonistius 
Lepomis. 
APHREDODERIDdJ. Pirate perches. 
Subfamily AphrEdodemnjE. 
Aphredoderus sayanus Gilliams. Houston, Tex. (Evermann). Scales about 2^3 mm. long and 
broad, parallel-sided, with four very deep basal plications, the basal margin strongly scalloped; 
apical margin with a row of about 25 to 30 very long teeth, free only at apex; nucleus subapical; 
nuclear area with fine labyrinthiform markings; lateral circuli widely spaced, but with rudiments 
of others between them. This is very different from Percopsis, though agreeing in having a single 
row of apical teeth. Except for the ctenoid feature, there is a rather strong resemblance to Lucius. In 
the character of the ctenoid fringe, there is much resemblance to the ctenoid Characinidas of Africa. 
The ctenoid structures in Percopsis and Aphredoderus are quite different from those in Perea, but a 
fairly intermediate type occurs in Pomacanthus arcuatus (Linnseus). 
Subfamily Trichophanin^. 
Trichophanes foliarum Cope, from the Miocene shales of Florissant, Colo., has been discussed and 
figmed in the American Natmalist, volume XLii, pages 570-574. It is there regarded as the type of a dis- 
tinct family, but I now believe that it deserves only subfamily rank. The scales have a subapical 
