No. 501] THE FLORISSANT EXPEDITION OF 1908 573 



tically complete specimens, herewith illustrated. These 

 reveal many characters not visible in the type, and em- 

 phasize the Percopsis-like tendencies. In Jordan and 

 Evermann's "Fishes of North and Middle America," 

 plates CXXI and CXXII, are given excellent figures of 

 Percopsis, Columbia and Aphredoderus. Our Tricho- 

 phanes agrees with Aphredoderus in the thick (deep) 

 caudal peduncle, the projecting lower jaw, and the scaly 

 sides of the head. The dorsal fin, as in Aphredoderus, 

 has three spines, the first very short, the third long (about 

 12.5 mm.), the second intermediate. The anal, as in 



Fig. 3. Caudal fin of Amia. 



Aphredoderus and Columbia, but not as in Percopsis, has 

 two spines, one long, the other short; the longer spine is 

 nearly straight, as in Aphredoderus. The shape of the 

 dorsal is very much like that of Percopsis, not very like 

 that of Aphredoderus; while the forked caudal is very 

 unlike that of the latter genus, but rather closely resem- 

 bles that of Columbia. There is no adipose fin (it is 

 present in Percopsicte) ; the ventrals are inserted about 

 5 mm. posterior to the bases of the pectorals, and the 

 same distance anterior to the level of the beginning of 

 the dorsal. In the last character the fish is nearly in- 

 termediate between Aphredoderus and the Percopsidae. 



Trichophanes should apparently be taken as typical of 

 a family Trichophanidffi, falling in the Xenarchi, and 



