THE ANCESTRY OF THE VERTEBRATES 



66l 



chordates, but, unfortunately, do not connect it with any phylum of 

 invertebrates. 



That the so-called notochord of balanoglossus is a true chorda dorsalis 

 like that of other chordates has been doubted. But Ritter (1900), on 

 the basis of his studies on Harrimania, a close relative of balanoglossus, 

 says that "there can be no serious doubt that we have in this organ the 

 immediate genetic forerunner of the vertebrate notochord." Its peculiar- 

 ities, in Ritter's opinion, do not warrant giving it another name, as has 

 been suggested. In Harrimania, two regions of the notochord are dis- 

 tinguishable, an anterior nuchal and a posterior esophageal. The poste- 



NOTOCHORD^ 



-DORSAL ARTERY 



^^ PHARYNX^^^ 



METAPLEURAL FOLD, 



, NERVE CORD. 



/^ 1/ GONAD--' I I ^ 



w \c 



•^prBIRRANPHIAL CAVITY. ^^^ 



PERIBRANCHIAL CAVITY. 



GILL APERTURE'. 



I VENTRAL ARTERY. 



GILL APERTURE 



VENTRAL ARTERY^' 



DIGESTIVE GROOVE. 



A. AMPHIOXUS. B.PTYCHODERA. 



Fig. 535. — Diagrams illustrating divergent methods by which the peribranchial 

 cavity is formed. In Amphioxus {A) the pleural folds are separated from the pharynx 

 by paired folds which extend dorsally from the ventral side. In the hemichordate 

 Ptychodera {B), on the other hand, the paired folds begin to form at the dorsal side of 

 the worm and extend ventrally. The peribranchial cavity in urochordates arises in a 

 similar manner. As frequently happens in animals, a similar end-result is attained by 

 divergent means. (Redrawn after Gaskell.) 



rior is embryonic and transient in other genera of hemichordates, but 

 persists in Harrimania, which must therefore be regarded as a more 

 primitive type. In the possession of a lumen, the notochord of balano- 

 glossus is peculiar; but in some chorda te embryos as in Harrimania the 

 notochord arise as a grooved outpocketing of the median dorsal wall of 

 the endoderm. (Fig. 534) 



While the connexion of balanoglossus with lower vertebrates is obscure, 

 the afhnities of its tornaria larva with the larvae of echinoderms and 

 with such primitive types as Cephalodiscus and Rhabdopleura are sug- 

 gestive of relationship with lower invertebrates. 



APPENDICULARIA THEORY 



Following Kowalewsky's (1868) researches upon the embryology of 

 urochordates and amphioxus, it was supposed that the long-sought 

 ancestor of vertebrates had been found, and the tunicate larva was pro- 



