EVOLUTION OF THE VERTEBRATES 



145 



Amphioxus theory, afford a very satisfactory hypothetical con- 

 nection of vertebrates and invertebrates. 



The arthropod theory is at fault chiefly in tracing the vertebrates 

 to invertebrates which were themselves highly specialized and 

 has not been given general support. The fishes through which the 

 connection is traced are looked upon as specialized types and such 

 resemblance as they show with fossil arthropods may therefore 

 be due to convergence. 



Hypothetical Ancestors. It is generally agreed that the ances- 

 tors of the vertebrates must have been free-swimming, active, 



Fig. 82. — Am-phioxus lanceolatus. (PYom Newman, after Wille3^) 



aquatic animals. They are equally certain to have been bilaterally 

 symmetrical with well developed heads, and definitelj^ metameric 

 in structure. They must also have possessed the three funda- 

 mental characters of the chordates, viz., tubular dorsal nervous 

 system, notochord, and pharyngeal clefts. 



The fishes are recognized as the most primitive of the true 

 vertebrates, and such an ancestor could not have been very dif- 

 ferent from the typical fishes. The existing lancelets {Branchios- 

 toma or Amphioxus) of all animals below the vertebrates best fit 



