638 INVEBTEBBATE MOBPHOLOGT. 



of the simple Tunicates (see text-books of embryology) are so very similar 

 to those of AmpMoxus that it must be concluded that the evolution of the 

 Ilrochorda and Cephalochorda proceeded for some distance along similar 

 lines, and the general af&nities of the Protochordata may possibly be indi- 

 cated by a scheme thus : .^^ ^ 



Vertebrata 

 Cephalochorda 



Hemichorda 



Urochorda 



Ancestral Protochordata 



Taking the larval Tunicates as a basis for comparison, we find as 

 features common to them and Amphioxus a dorsal nervous system aris- 

 ing as an invagination of the ectoderm and extending the entire length 

 of the body ; in the anterior portion the lumen of the nerve-cord expands 

 to form a brain which in Amphioxus opens in early stages to the exterior 

 and in the Tunicates into the anterior portion of the branchial sac, i.e., 

 the ectodermal portion, the canal of communication in the latter forms 

 losing in later stages its connection with the brain and forming the sub- 

 neural gland. An atrial cavity occurs in both, which, though arising in a 

 somewliat different manner in the two groups, nevertheless seems quite 

 homologous, and homologies have also been pointed out between the 

 primary gill-slits. The increased number of stigmata and their arrange- 

 ment in the Tunicates is a secondary character resulting probably from the 

 sessile existence ; and the development of the test and the limitation of the 

 notochord to the tail are also probably secondary characters. The resem- 

 blances are important ones, and when taken into consideration with the 

 embryonic development point very strongly to a close affinity. 



As regards the relationships of the various groups of Urochorda to one 

 another considerable difference of opinion exists. The Appendicularians, 

 which at first sight seem to be the most primitive of all the orders, present 

 certain remarkable peculiarities, such as the separate openings of the atrial 

 cavities and the anus, aud some authors are inclined to regard them not as 

 primitive forms, but as sexuall.y-mature larvae of sessile forms in which a 

 test had already developed and degeneration far advanced. As regards 

 the remaining forms the simple Ascidians seem to be the most primitive, 

 the composite forms being derived from them by the acquisition of non- 

 sexual reproduction. The composite forms, however, seem really to 

 represent several groups originating independently, all the members not 

 having descended from an ancestral simple form, but some from one 

 ancestor and others from another, and so on. The Thaliacea, finally, have 



