340 BULLETIN OF THE 



the ordinary Ascidians. Bronn (Klassen und Ordnungen) gives a 

 somewhat similar series, arranged in the same order, but beginning 

 with a Polyzoon and ending with a Lamellibranch and a Brachiopod. 

 It is, of course, unnecessary to enlarge, in this place, upon the fact 

 that the features which have been supposed to unite the Tunicata 

 with the Polyzoa, Brachiopods, aud Lamellibranchs are superficial and 

 absolutely without scientific value, as this is now recognized by ull 

 who are familiar with what is known of the embryology of these 

 various animals. 



That the position assigned to Salpa, in the series above referred to, 

 rests upon a false conception of the nature of the " gill " and its rela- 

 tions to the branchial sac in the ordinary Tunicates, has never been 

 pointed out, but the history of the formation of this structure, as it 

 has been described in this paper, shows that it is not a rudimentary 

 but a specialized form of the branchial sac ; and this, as well as all the 

 other peculiarities of Salpa, seem to be special adaptations to its mode 

 of life, and seem to show that it occupies a very high position among 

 the Tunicata. 



So little is known of the life history of Appendicularia and Doliolum 

 that their relation to the remaining members of the group cannot be 

 stated with any certainty ; but as far as we now know, Appendicularia 

 seems to be an adult representative of the " tadpole larva " stage, and 

 must be regarded, for the present at least, as a very low and embry- 

 onic form, while Doliolum appears to stand between Salpa and the 

 ordinary Tunicata as a transitional form.* 



Salpa lacks the " tadpole larva " stage, and this is what we should 

 expect in a Tunicate which had become adapted to a locomotive life. 

 The power of locomotion during the earlier stages is very necessary 

 to those animals which subsequently become fixed, and we should ex- 

 pect, according to the theory of evolution, that, wherever these fixed 

 animals are descended from a free ancestor, they should retain, dur- 

 ing the early stages, the free locomotive form of their remote prede- 

 cessors. 



* During my work on Salpa specimens of Appendicularia were frequently met with, 

 hut no thorough study of them was made, owing to lack of time. Although a constant 

 search for Doliolum was kept up, only two dead specimens were met with during the 

 summer, so that no opportunity for filling the gaps in our knowledge of this form was 

 afforded. 



