On Fossil Fishes from the Karroo System, and some General Considerations on the Bouy Fishes of the Triassic Period

Summary. The foregoing discussion on this group of Palsoniscid derivatives reveals three main points: That the two families owe their similarity to a Palsoniscid origin and to the limited nature of the evolutionary changes in which they have subsequently been involved. In consideration of the distinct differences in the structure of the skull, the lack of intermediate forms, and their co‐existence side by side throughout Triassic time, it is probable that the two families arose from separate Palaoniscid stocks. Their many resemblances to the Palaeoniscids are genuine, and denote a true relationship. The resemblances to the Semionotidae are, on the other hand, clearly due to convergence, for the structure of the head in Semionotids is much modified from that of these or any of the older groups. Stensio's suggestions that the Semionotidae may have been derived from the Catopterid‐Perleidid group is very highly improbable, for the Semionotids had already arisen as a separate Offshoot from the Palsoniscids in Mid‐Permian time, apparently long before either Catopterids or Perleidids had come into existence.