EVOLUTION AND GEOLOGICAL HISTORY 171 



question arises whether they can be distinguished from reptiles, 

 especially from the reptiles of the same or slightly later 

 periods. It is impossible to avoid the conclusion that the 

 existing Amphibia as well as the earliest reptiles were derived 

 from the Stegocephali : the latter are distinguished by the pos- 

 session of two occipital condyles or none at all, by the 

 structure of the vertebrae, and by the primitive characters of 

 the skeleton to which that of the Apoda and Urodela is evi- 

 dently allied. It is usually stated that the whole of the dorsal 

 surface of the skull is covered or roofed over by dermal bones, 

 which means that in these primitive Amphibia the flat or 

 membrane-bones of the skull were still in the condition of 

 dermal scutes, as in the primitive bony fishes such as the African 

 fish Polypterus and the lung-fishes. Dermal bones, or scales, 

 were also present in many types of the group on the body, either 

 on all parts or only on the lower surface, but some were desti- 

 tute of such structures. The fossil skeletons naturally do not 

 throw much light on the condition of the gills or lungs, but in 

 the Branchiosauri, one of the sub-orders, representatives of 

 which are found in the Carboniferous and Permian formations, 

 the young specimens show very distinct gill-arches with 

 numerous nodules and denticles on them ; there can be little 

 doubt that these arches carried functional gills like those of 

 the aquatic larvae of existing Amphibia. Branchiosanrus sa- 

 lamandroides is found abundantly in the Permian beds of 

 Europe and specimens are found of almost every size and 

 stage of development from larvae of three-quarters of an inch 

 to adults of two and three-quarter inches. Gill-arches and 

 gills seem to be absent in the adult so that in all probability 

 this form at least went through the characteristic Amphibian 

 metamorphosis. In the Carboniferous strata are found skele- 

 tons of another sub-order in which the body was long and 

 snake-like, without any limbs ; in these gill-arches have been 

 recognised behind the head and these carry skeletal rods 

 supposed to have been the supports of external gills which 

 were retained throughout life ; it has been suggested that these 

 were the ancestors of the existing Apoda, but no connecting 

 forms from intermediate periods are known. In the Carboni- 

 ferous are also found Keraterpeton and Uvocordylus, animals 

 shaped like a newt, with a ventral armour, and about a foot in 



