xiii PHYLUM CHORDATA 595 



the highly specialised classes referred to. If, for instance, the 

 lower Craniata sprang either from a Chaetopod-like or from a 

 Limulus-like ancestor, Amphioxus and the Tunicates must either 

 have no connection at all with Vertebrates, or must have under- 

 gone a quite inconceivable amount of degeneration. 



There appears, in fact, to be a good deal of evidence supporting 

 the view that the first developed Chordates may not have had any 

 such pronounced metamerism as Annulates and Arthropods possess, 

 and it becomes possible to look for the immediate ancestral form 

 among less highly organised Invertebrate groups. But on this 

 matter, though there is abundant scope for ingenious speculation, 

 no finality can be said to have been reached, and the ancestry of 

 Vertebrates still remains an unsolved problem. 



With regard to the higher classes, Amphibians may be held to 

 have arisen from a Fish-type allied to the Dipnoi, the resem- 

 blances of which to the Amphibia are so great as to lead some 

 authors to place them in a distinct class intermediate between 

 Fishes and Amphibia. The chief difficulty in this case and it 

 is a serious one is the derivation of the pentadactyle limb 

 from the Fish's fin, a transformation of which no satisfactory 

 explanation is at present offered either by anatomy, embryology, 

 or paleontology. 



Reptiles may be considered to have arisen from a generalised 

 Amphibian stock, but there is no direct evidence on this point ; 

 and, apart from purely theoretical considerations, there is nothing 

 to show how or why gills vanished so completely as to leave no 

 trace of their existence apart from the branchial clefts, or by what 

 steps the allantoic bladder became precociously enlarged into an 

 embryonic respiratory organ. The precise mode in which the pro- 

 tecting amnion arose is also very doubtful, though from theoretical 

 considerations it has been argued that its development in the 

 Hedgehog (p. 574) indicates a more primitive condition than 

 obtains in the other Mammalia or even in Sauropsida. 



Birds appear to be undoubtedly derived from true Reptiles, 

 although nothing is known of the actual ancestral form.' In spite 

 of the enormous adaptive differences between the warm-blooded, 

 feathered, bipedal Bird and the cold-blooded, scaly, quadrupedal 

 Reptile, the connection between the two is far closer than between 

 any other two vertebrate classes. 



Mammals also appear to have had a reptilian origin : the 

 numerous reptilian characters of the Monotremata certainly point 

 in this direction, and the reproductive processes of that group help 

 us to understand the stages by which the large-yolked egg of the 

 ancestral form, with the embryo developed outside the body, 

 may have given place to the secondarily alecithal egg of the 

 typical Mammal, giving rise to a foetus developed within 



