xiii. 3 AQUATIC AMPHIBIA 361 



bians whose affinities are less certain. They have in common a reduc- 

 tion of ossification in general and in particular in the centra, which 

 seem not to be formed from separate cartilaginous elements as in 

 labyrinthodonts but as thin continuous sheets of bone. Some of these 

 animals classed as Phyllospondyli or 'Branchiosaurs' were almost 

 certainly larval Rhachitomi; the external gills can be recognized and 

 stages found connecting them with known adults of that group. It is 

 necessary, however, to retain the order for the present. 



Other early aquatic amphibia are less easy to classify and are 

 grouped for convenience as an order Lepospondyli, all having verte- 

 brae composed of a single piece, and a continuous notochord. They 

 show, however, at least three distinct lines, probably separate off- 

 shoots from the main labyrinthodont stock. *Dolichosoma and other 

 forms from the Carboniferous were like snakes and had lost the limbs. 

 *Diplocaulus from the Permian possessed remarkable horned skulls. 

 These creatures with broad flat heads and upward-looking eyes and 

 small limbs were presumably bottom-dwellers and the development 

 parallels that of the Stereospondyli. A third aberrant group placed 

 here, the Microsaurs, such as *Microbrachis, were animals with long 

 bodies and small limbs, presumably aquatic, but showing many 

 similarities to the reptiles in the skull. 



The order Adelospondyli has been created for a further collection 

 of presumably secondarily aquatic amphibia such as *Lysorophus, in 

 which the neural arch and centrum are not fused but articulate by 

 means of a jagged suture. The skull shows reduction and variation 

 of the bones, and for this and other reasons it has been suggested that 

 the urodeles may have arisen from an adelospondylous line. 



The relationship of the modern amphibia to these palaeozoic 

 stegocephalians remains uncertain. The earliest anuran is *Proto- 

 batrachus of the Triassic, possessing ribs and a tail, but with elongated 

 ilia and an anuran type of skull. It is probably a larva in metamor- 

 phosis. *Miobatrachus of the Carboniferous is a form in which the 

 posterior portion of the skull is shortened and the temporal bones lost. 

 In other respects the skull is like that of a rhachitome and suggests 

 that the frogs diverged from the labyrinthodonts at this very early 

 stage. 



The urodeles can be traced back only to the Jurassic. It is often 

 suggested that both they and the Apoda have arisen from aquatic 

 Lepospondyli, such as the microsaurs, but there is no real evidence 

 of this. 



