200 HISTORICAL PALAEONTOLOGY. 



of the Middle Permian ; and the principal genera of the 

 Ganoids are Pateoniscus and Platysomus (fig. 137). 



The Amphibians of the Permian period belong principally 

 to the order of the Labyrinthodonts, which commenced to be 

 represented in the Carboniferous, and has a large development 

 in the Trias. Under the name, however, of Palaosiren Bcinerti, 

 Professor Geinitz has described an Amphibian from the Lower 

 Permian of Germany, which he believes to be most nearly 

 allied to the existing " Mud-eel " (Siren, lacertina) of North 

 America, and therefore to be related to the Newts and Sala- 

 manders (Urodela). 



Finally, we meet in the Permian deposits with the first un- 

 doubted remains of true Reptiles. These are distinguished, as 

 a class, from the Amphibians, by the fact that they are air- 

 breathers throughout the whole of their life, and therefore are 

 at no time provided with gills; whilst they are exempt from 

 that metamorphosis which all the Amphibia undergo in early 

 life, consequent upon their transition from an aquatic to a 

 more or less purely aerial mode of respiration. Their skel- 

 eton is well ossified ; they usually have horny or bony plates, 

 singly or in combination, developed in the skin; and their 

 limbs (when present) are never either in the form of fins or 

 wings, though sometimes capable of acting in either of these 

 capacities, and liable to great modifications of form and struc- 

 ture. Though there can be no doubt whatever as to the occur- 

 rence^ genuine Reptiles in deposits of unquestionable Per- 

 mian age, there is still uncertainty as to the precise number 

 of types which may have existed at this period. This uncer- 

 tainty arises partly from the difficulty of deciding in all cases 

 whether a given bone be truely Labyrinthodont or Reptilian, 

 but more especially from the confusion which exists at pres- 

 ent between the Permian and the overlying Triassic deposits. 

 Thus there are various deposits in different regions which 

 have yielded the remains of Reptiles, and which cannot in 

 the meanwhile be definitely referred either to the Permian 

 series or to the Trias by clear stratigraphical or palaeonto- 

 logical evidence. AH that can be done in such cases is to be 

 guided by the characters of the Reptiles themselves, and to 

 judge by their affinities to remains from known Triassic or Per- 

 mian rocks to which of these formations the beds containing 

 them should be referred ; but it is obvious that this method 

 of procedure is seriously liable to lead to error. In accor- 

 dance, however, with this, the only available mode of deter- 

 mination in some cases, the remains of Thecodontosaurus and 

 Paltzosaunis discovered in the dolomitic conglomerates near 



