=Erismatopteridæ.=—Here should perhaps be placed the family of _Erismatopteridæ_, represented by _Erismatopterus levatus_ and other species of the Green River Eocene shales. In _Erismatopterus_ the short dorsal has two or three spines, there are two or three spines in the anal, and the abdominal ventrals are opposite the dorsal. Allied to _Erismatopterus_ is _Amphiplaga_ of the same deposits. We cannot, however, feel sure that these extinct fragments, however well preserved, belonged to fishes having an adipose fin. Among spiny-rayed fishes the _Percopsidæ_ alone retain this character, and the real affinities of _Erismatopterus_ may be with _Aphredoderidæ_ and other percoid forms. The relations of the extinct family of _Asineopidæ_ are also still uncertain. This group comprises fresh-water fishes said to be allied to the _Aphredoderidæ_, but with the pelvic bones not forked. _Asineops pauciradiata_, _squamifrons_ and _viridensis_ are described from the Green River shales. With _Erismatopterus_ all these fishes may belong to the suborder of _Salmopercæ_, but, as above stated, the possession of the adipose fin, the most characteristic trait of the _Salmopercæ_, cannot be verified in the fossil remains. [Illustration: FIG. 194.—Shoulder-girdle of the Opah, _Lampris guttatus_ (Brünnich), showing the enlarged infraclavicle. (After Boulenger.) ] =Suborder Selenichthyes, the Opahs: Lamprididæ.=—We may bring together as constituting another suborder certain forms of uncertain relationship, but which seem to be transitional between deep-bodied extinct Ganoids and the forms allied to _Platax_, _Zeus_, and _Antigonia_. The name of _Selenichthyes_ (σηλήνη, moon; ἰχθύς, fish) is suggested by Boulenger for the group of opahs, or moonfishes. These are characterized by the highly compressed body, the great development of a large hypocoracoid, and especially by the structure of the ventral fins, which are composed of about fifteen rays instead of the one spine and five rays characteristic of the specialized perch-like fishes. The living forms of this type are further characterized by the partial or total absence of the spinous dorsal, by the small oblique mouth, and the prominence of the ventral curve of the body. A thorough study of the osteology of these forms living and fossil will be necessary before the group can be properly defined. The large bone above mentioned was at first considered by Boulenger as the interclavicle or infraclavicle, the hypocoracoid being regarded by him as displaced, lying with the actinosts. But it is certain, from the studies of Mr. Starks, that this bone is the real hypocoracoid, which in this case is simply exaggerated in size, but placed as in ordinary fishes.