THEORY OF EVOLUTION 485 



be remembered from the previous study of Balanoglossus, it possesses 

 gill slits, a support in the base of the proboscis which may be 

 homologous to the notochord, and four longitudinal nerve cords of 

 which the dorsal is the most highly developed. The above proto- 

 chordate relations are rather generally conceded, but there is much 

 less agreement concerning their origin, and several theories have 

 arisen of which the following are important. 



Annelid Theory. — The segmental condition of this group, the re- 

 lationship of the digestive system to other circulatory and nervous 

 systems, and presence of the coelom with related nephridia, present 

 a close comparison to what is found in the embryonic development 

 of the vertebrates. It has been suggested by some scholars that by 

 inverting the body of the nonchordate annelid, the fundamental 

 systems are brought to resemble their relative locations in verte- 

 brates. A fibrous cord has been found in some groups of the an- 

 nelids, and this structure is held to take the place of a notochord in 

 function and position. These fibers are found just dorsal to the chain 

 of ganglia in the annelids. 



Arachnid Theory. — Such forms as scorpions, Limulus, and other 

 arachnids have been favorably compared to vertebrates. By com- 

 parison of these arachnids with the extinct, fishlike ostracoderms, 

 an elaborate theory of the possible origin of the vertebrates from 

 this ancestry is derived. 



Echinoderm Theory. — This theory of vertebrate descent goes again 

 to the Balanoglossus. The developing egg of this animal becomes 

 a larva known as Tornaria (Fig. 47), which floats in marine waters, 

 has bilateral symmetry, is almost transparent, and possesses bands 

 of cilia used in locomotion. This larva is almost exactly like that of 

 the starfish and other echinoderms which live in the same habitat 

 (Fig. 140). The close correspondence of features of these two groups 

 of larvae has suggested the conclusion that these two types of ani- 

 mals have descended from a common ancestor which was similar to 

 these larvae. The line of descent of one branch of this stock has 

 presumably passed through Balanoglossus, Tunicata, Amphioxus, and 

 Vertebrata. The nonchordate ancestors are not yet conclusively 

 determined, but the foregoing theories suggest the thinking and evi- 

 dence along that line. 



Within the class Vertebrata the relations are somewhat more evi- 

 dent, but the phylogenetic sequence is rather obscure at some points. 



