122 THE VOYAGE OF H.M.S. CHALLENGER 



important ones/ and of a large proportion of the known species. I state all these 

 details merely to show that any speculations which follow, whether they be correct 

 or not, are at least not unfounded, but are based upon an extensive series of 

 observations.' 



In rcQ-ard to the relations of the Tunicata as a whole, I still incline to the view 

 which was stated first by Balfour in 1881," and to which I have adhered in several 

 papers* published since, viz., that the Tunicata are to be regarded as a degenerate 

 ofi'shoot from the Protochordata, an early group of common ancestors of the Tunicata, 

 AmpMoxus and the Vertebrata. This position was adopted by E. van Beneden and 

 Julin in their Recherches sur la Morphologic des Tuniciers,'' and supported by 

 a strong body of embryological evidence which goes to demonstrate (1) the funda- 

 mental agreement between the development of the larval Ascidian, which may be 

 taken as the nearest form we know to the ancestral Tunicate, and Amj^hioxus, which 

 is the nearest representative we have of the ancestral Chordata; and (2) the independ- 

 ence or considerable difference of some organs or systems in the body of the Ascidian 

 from the corresponding parts in the Vertebrata. 



This view of the origin of the Tunicata is controverted by Dohrn," who regards the 

 group as having degenerated, not from the Protochordata, but from the early Verte- 

 brata, such as the lower fishes. This matter is still decidedly open to and is now under 

 discussion ;' but as the Challenger anatomical observations do not furnish any fresh 

 evidence for either side, it is unnecessary to discuss it further here. 



In the phylogenetic diagram (p. 120) the Prototunicata are show^n as arising 

 ;is an offshoot from the Protochordata not far from the point where the ancestors of 

 Amjyhioxus left the main line. If, on the other hand, the vertebrate affinities of the 

 group were adopted, the point of origin would have to be placed farther up the 

 Chorda,te branch, and our conception of the early Prototunicata would be somewhat 

 modified ; but the remainder of this phylogenetic scheme from the later Proto- 

 tunicata, now represented by the tailed larval Ascidian, onwards would remain 

 unaffected by the change. 



From a phylogenetic point of view the Tunicata are especially interesting — (1) on 



' The only genera I have not had an opportunity of examining for myself are a few unimportant recently formed 

 ones, such as Styelopsls and Paramol(jiilii, which are closely related to older well-known genera, and present no points 

 of fundamental importance. 



- It would take uj) too much space to mention all the details of structure and development upon which the con- 

 clusions given in the following pages depend. The more important points are, however, discussed, and reference is made 

 in the footnotes to the original works in which the structure of the various forms is described. 



' Comparative Embryology, vol. ii. p. 271. 



* A Phylogenetic Classification of iVnimals, p. 58, London ; Macmillan, 1885 ; this Report, Part II., p. 387, 1886 ; On 

 the Phylogeny of the Tunicata, Proc. Roy. Soc. Edin., vol. xiii. p. 444, 1 88C ; and Enoy. Brit., 9th ed., article " Tunicata.'' 



' Archives de Biolvgie, torn. vi. p. 459, 1887. 



^ Studien zur Urgeschichte d. ^Virbelthierkorpers, Mitth. d. zool. Stat. Neapcl, vol. vi. 1885. 



' See Dohrn's Studien, etc., Mitth. d. zool. Stat. Neapel, vol. vii. 1887 ; and Zoohgischer Anzcigcr, Jahrg. x. pp. 

 407 and 433, 1887. 



