56 



which * Professor Huxley considers the passage from the fishe^ 

 to the reptiles took j)lace. All the families of this well-defiiieil 

 sub-order are characterised by the possession of two dorsal fins, 

 and by lobate paired fins having a central axis or stem covered 

 with scales like the body walls, and surrounded by a fringe of fin 

 rays. Jugular plates always replace the branchiostegal rays, and 

 the scales are either rhomboidal or cycloidal. The families 

 Scmrod'ipferiui, GUjiitodipterini, and Phanerophminl are restricted 

 to the Palaeozoic rocks. The Caelacanthini range from the Car 

 boniferous to the Chalk and the PoJypterini, comprising only the 

 living Pobjpteriis and Calamoichthys of Africa alone rejiresent this 

 numerous race of fishes at the jiresent day. The genus 

 Pulypterus is remarkable for the unique arrangement of its sub- 

 divided dorsal fin, and by the possession of a double cellular air 

 bladder, which most nearly ajjproximates to the true lungs of the 

 Dipnoi. It has least structural affinities with the Ccdaainths, its 

 nearest allies in time, and is most closely zoologically related to 

 the rhomboidal scaled Saurodipterines of the Devonian, from 

 which it is separated by an enoimous gulf of geological time, as 

 no intermediate links have been discovered. 



In the notochordal Phanenqjleurini we find forms which most 

 closely resemble the acutely lobate finned Lepidosiren. The 

 shape of the body, number, position, and structure of the fins, 

 and all the elements of the internal skeleton, exactly foreshadoAv 

 those of the mud fishes. Like them Phanewpleuyon was covered 

 with thin cycloidal scales, through which the long and well 

 ossified ribs show so jjlainly in the fossil state as to suggest the 

 name of the genus. The dentition, however, differs from that of 

 Ceratodus and Lepidosiren, being composed of a row of short 

 conical teeth in each jaAv, and in the absence of the grooved 

 dental jjlates so characteristic of the true dipnoi it is uncertain 

 whether this family can be associated with the other members of 



* Decade X of the Memoirs of the Geological Survey, 1861 (Classification 

 of Devonian Fish). 



