EVOLUTION AND GEOLOGICAL HISTORY 175 



sense-organs are present in newts during the breeding season 

 when they live in the water. 



Dr. Gadow reconstructs the stages in the evolution of Am- 

 phibia as follows : — 



(1) Terrestrial with two pairs of limbs with five toes on each, 

 breathing by lungs only, with five pairs of gill-arches, which 

 during embryonic life perhaps carried internal gills; with or 

 without several pairs of gill-clefts. 



(2) External gills were developed in the embryo, i.e. before 

 hatching, and were afterwards retained later during larval life ; 

 these external gills superseded the internal gills of which there 

 are now no traces in Urodela or Anura (p. 159). 



(3) Some Urodeles, prolonging the aquatic life, retained and 

 enlarged the external gills into more or less permanent organs. 



(4) Some Urodela, e.g. Salamandra atra, have become en- 

 tirely terrestrial, the larval metamorphosis being passed through 

 in the maternal oviduct. The possession of unusually long 

 external gills by this species and by Apoda indicates that these 

 organs are essentially embryonic, not larval features. 



The present writer is unable to agree with these conclusions 

 entirely. The presence of external gills in the larvae of Polyp- 

 terus and the lung-fishes among fishes, from which the Am- 

 phibia were derived, indicates that these organs were probably 

 already present in the larval stage of the earliest Amphibia. 

 The special elongation of the external gills in the embryos of 

 Apoda and in those of Salamandra while still within the maternal 

 uterus, may well be regarded as an adaptation to the needs of em- 

 bryonic life, but their original development in lung-fishes and Am- 

 phibia is evidently a larval feature and was very probably due to 

 the turbidity of the water in which the larvae lived, which made 

 the ordinary mode of respiration in fishes difficult. On the other 

 hand, there can be no doubt that the transformation of the 

 paired fins of the fish into the terrestrial limbs of the first Am- 

 phibia was due to the use of the limbs for supporting the body 

 and moving on land ; mere use of the limbs on the ground in 

 shallow water would scarcely have 'sufficed to bring about so 

 great a structural change, and there is no evidence in the fossil 

 Labyrinthodonts of forms with well-developed terrestrial limbs 

 having gills or gill-clefts in the adult state. We know from the 

 actual structure of the existing lung-fishes that the lungs were 



