664 



COMPARATIVE ANATOMY 



ancestors and that urochordates are degenerate vertebrates. Brooks 

 denies the truth of both of these assumptions. 



The objection most frequently raised against the appendicularia 

 theory is that tunicates are chordates, and that this theory does not 

 carry us beyond the Hmit of the phylum. The purpose of a theory of 

 vertebrate ancestry, however, is to reveal the non-chordate predecessors of 

 chordates. But there appear to be no invertebrates which can be con- 



LATERAL TRUNK MUSCLES 



SPINAL CORD 

 NOTOCHORD 



DORSAL AORTA- 



PRECARDINALV 



EPIBRANCHIALGROOVE ||-^:S?£ 



GILL LAMELLAE 



CARTILAGE BAR 

 GILL-RODS 



TRANSVERSE MUSCLE 

 VENTRAL AORTA 

 PERIBRANCHIAL CAVITY 



HYPOBRANCHIAL MUSCLE 

 METAPLEURAL FOLD 



A. AMPHIOXUS B. AMMOCOETES. 



Fig. 539. — Cross sections of A, Amphioxus and B, Ammocoetes (larval Petromyzon) 

 through the pharyngeal region showing their fundamental resemblance. Their simi- 

 larities even in details of histological structure seem too great to be the result of con- 

 vergence of genetically unrelated forms. Just as the search for the most primitive 

 forms of vertebrates leads from elasmobranchs to cyclostomes, so in turn Amphioxus 

 shows characters similar to, but simpler than, those of vertebrates. It is therefore not 

 surprising that most morphologists place amphioxus near the main line of vertebrate 

 ancestry. 



sidered the immediate predecessors of chordates — no forms transitional 

 between chordates and the much simpler coelenterates. 



CONCLUSION 



A survey of these various attempts to solve the problem of the ancestry 

 of vertebrates seems to suggest that the most promising clue to the solu- 

 tion is to be found in the lower chordates. Cyclostomes, more especially 

 their larval stages, lead us in the direction of Amphioxus, which "if it 

 hadn't existed, would have had to be invented." The similarity of 

 amphioxus embryos to larval tunicates strongly suggests their common 

 origin. Since this seems the most reasonable interpretation of the facts, 

 we may conclude that metamerism has been attained de novo by chor- 



