TYPE PROTOGHOBDATA. 617 



of them the branchial blood-vessels form a small plexus which 

 may be regarded as a glomerulus such as occurs in connec- 

 tion with the urinary tubules of the Vertebrate kidney. 



Amphioxus is bisexual. The reproductive organs, ovaries 

 or testes (Figs. 278 and 279, g), occur in the epipleural folds, 

 and are arranged metamerically, extending in A. lanceolatus 

 from the fifteenth to the thirty-fifth or thirty-sixth metamere. 

 They lie at the level of the junction of the lateral longitudinal 

 and the ventral transverse muscles, and are contained within 

 a coelomic cavity which is a portion of the original myocoels 

 of the segments to which the reproductive masses belong. 

 They lie on the inner surface of the atrial folds, and are 

 covered on their inner surfaces by the thin bodj'^-wall, through 

 which the reproductive elements break when mature, falling 

 into the atrium and thence passing to the exterior through 

 the atrial pore. There are no reproductive ducts. 



The Affinities of Amphioxus. — The Oephalochorda have usually been 

 considered the most primitive "Vertebrates, — as representing, in other words, 

 in a more or less modified condition the stem from which the Vertebrata 

 have descended, — and there are many points of similarity between Amphi- 

 oxus and the larval Lampreys {Ammocetes) by which such a view may be 

 supported. The character and origin of the notochord and nervous 

 system, the arrangement of the nerves of the latter, the lateral muscle- 

 plates, the occurrence of an hepatic portal circulation and the character of 

 the early stages of development are all points of similarity which seem 

 explicable only on the supposition of a somewhat close aflSnity. 



On the other hand, resemblances to the Enteropneusta are but slightly 

 less marked. It seems hardly possible that the marvellous similarities in 

 the arrangement of the branchial skeleton should have been acquired in- 

 dependently in the two forms, and the atrial folds of Amphioxus may be 

 regarded as the more extensively-developed atrial folds of Bdlanoglossus. 

 Amphioxus is undoubtedly much more advanced along the Vertebrate line 

 than Balanoglossus, and both are probably more or less widely divergent 

 from the direct line, but the general similarities, such as the occurrence of 

 branchial slits, of an endodermal notochord, and of a dorsally-placed 

 central nervous system, which have led to the association of both forms in 

 the type Protochordata, speak strongly for a community of descent. 



The principal difiBculty in the way of the acceptance of Balanoglossus 

 as the representative of the ancestral forms from which the Vertebrates 

 are descended seems to lie in the supposed necessity for deriving highly- 

 organized metameric forms, such as the Vertebrates, from more lowly but 

 also metameric ancestors, and consequently most authors have sought for 



