CLASSIFICATION 463 



garding Lancelets as derivable from higher Chordates. Nor 

 does the development of Lancelets at all favour the idea that 

 they are degenerate. 



IV. There seems to be affinity — though not very close — 

 between Lancelets and Tunicates. In both cases the walls of 

 the pharynx are perforated by numerous gill-slits ; in both 

 there is a peribranchial cavity; in both there is an endostyle ; 

 there are several other resemblances all relating to remarkable 

 structural features. On the other hand, the Ascidians differ 

 from the Lancelets in a great many ways — the presence of 

 test, heart, and genital ducts ; the absence of segments, of 

 nephridia, and of any trace of ccelom in the adult. The general 

 habit of life is very different and the Ascidians are hermaphro- 

 dite. The early stages in development in Tunicates and Lan- 

 celets have much in common, but after these are past there is 

 great divergence. Now when we find very striking resem- 

 blances, e.g., in the pharynx, and very striking differences, e.g., 

 as regards segmentation, we are led to the conclusion that Lance- 

 lets and Tunicates, though not nearly related, are descended from 

 common ancestors, and that these ancestors diverged from the 

 main Chordate stock, which led on to fishes and other Verte- 

 brates. 



Classification 



There is not as yet unanimity in regard to the number of 

 genera of Lancelets ; we give Dr. Arthur Willey's version of 

 what is always a difficult matter. Two genera may be distin- 

 guished, one including those species with paired reproductive 

 organs, the other including those with unilateral reproductive 

 organs. For the first the most correct title seems to be Bran- 

 chiostoma, Costa, and it includes the sub-genera Amphioxus, 

 sensu stricto, Yarrell, and Dolichorhynchus , Willey. For the 

 second genus the name is Heteropleuron, Kirkaldy, and it in- 

 cludes as sub-genera Paramphioxus, Haeckel ; Epigonichthys, 

 Peters ; and Asymmetron, Andrews. 



An important general fact may be illustrated by the geo- 

 graphical distribution of Cephalochorda, namely, that old- 

 fashioned types often have a very wide representation. We 

 see this in some terrestrial types like Peripatus and its allies 

 and in some fresh-water types like the Dipnoi or mud-fishes; 



