A new specimen of Doryptems hoffmanni Germar, from the Marl Slate of England, is described and a new account of the anatomy of the fish given. Modification of the excellent earlier reconstruction by Gill is necessary in the skull-roof and jaws and the opercular apparatus, and the dorsal fin is longer than hitherto known. A new explanation of the working of the jaws is given, and the lack of ossification in the head is shown to be a corollary of the jaw-mechanism. The reduction of the opercular apparatus, and presumably of the gills, is considered in relation to possible air-gulping habits. The body-form and fin-structure are reviewed in relation to the methods of swimming, The distribution of heavy bones and scales, the extremely compressed body the structure of the internal skeleton of the posterior trunk, and the jugular position of the pelvic fins are shown to be structural modifications connected with the swimming-necessities of a slow-swimming deep-bodied fish adapted for life in habitats not unlike those of coral-fishes. A review of the possible environments of the Marl Slate lagoons suggests that, Dorypterus and other deep-bodied forms lived in “forests” of subaqueous plants, probably algal, where their only predator can have been Cœlacanthus. It is probable that Dorypterus fed on soft algal growths, the jaws functioning in a scissor-like fashion. The relations of Dorypterus to other deep-bodied fishes are summarized, and it is concluded that it was probably derived from the same group of “Platysomid” fishes that gave rise to Bobasatrania; the Pycnodonts may be descended from the same radicle. Lekanichthys howsei Brough is a very incomplete Globulodus macrurus Ag., and has no relevance to the problem of Dorypterus.