STUDIES ON THE PROTOOHORDATA, 



349 



Mouth ventral. 

 No endostvle. 



Mouth dorsal. 

 Endostjle. 



Cephalodiycus. 

 Sessile. 

 U-shaped ali- 

 mentary canal. 

 One pair 

 gill-slits. 

 Buds. 



Balauoglossus. 

 Tree. 

 Strai;^ht ali- 

 mentary canal. 

 Many 

 gill-slits. 

 No buds. 



Ascidians. 



Sessile. 

 U-shaped ali- 

 mentary canal. 

 Three pairs 

 gill-slits. 

 Buds. 



Amphioxus. 

 Free. 

 Straight ali- 

 mentary canal. 

 Many 

 gill-slits. 

 No buds. 



The form which might appear at first sight to interfere with 

 the harmony of the above scheme is Appendicularia. Most 

 authors (see Seeliger, 31^ and Herdman, 18) agree in regard- 

 ing Appendicularia as representing more or less the ancestor 

 of the Ascidians. 



From what has been said above, however, about the 

 U-shaped alimentary canal, it would follow that Appendicu- 

 laria came into existence after the acquisition of this curved 

 intestine ; in other words, that Appendicularia is at any rate 

 less primitive than the most primitive sessile Ascidians. See- 

 liger agrees with van Beneden and Julin in regarding the 

 atrial cavities of the fixed Ascidians as homologous with the 

 two branchial tube-like apertures or atrial canals of Appendi- 

 cularia, and this is undoubtedly correct. In Appendicularia, 

 as is well known, there is a pair of tubular gill-slits but no 



