238 THE VOYAGE OF H.M.S. CHALLENGER. 



Tentacles simple, filiform. 



Intestine always extending beyond the branchial sac posteriorly. 

 Genitalia placed in the intestinal loop. Reproduction by gemmation is also usual, 

 and results in the formation of colonies. 



For a detailed account of my reasons for considering the Clavelinidse as Simple 

 Ascidians, I must refer to the second part of the Preliminary Report. 1 It will be 

 sufficient here to mention that in organisation the Clavelinida? are most closely allied to the 

 Ascidiidse ; that if we consider only the structure of a single individual, an Ecteinascidia 

 might be placed in the same genus w T ith a Ciona, and that the property of reproducing by 

 gemmation and forming colonies is really the only essential character which separates the 

 Clavelinidse from the Ascidiidse ; and finally, that this property of budding cannot 

 be considered as forming a strict line of demarcation, since all the apparatus necessary for 

 the process is present in a more or less developed condition in other Simple Ascidians, and 

 has been observed in species of Ascidia 2 and Ciona 3 to produce stolons similar to those 

 of a Clavelina. On account of their property of reproducing by gemmation the Claveliuidaj 

 must be considered as having a closer affinity to the Compound Ascidians than have any 

 others of the Simple Ascidians. They form a passage between such forms as Ciona in the 

 Ascidiidse and Diazona in the Compound Ascidians. 



The Challenger genus Ecteinascidia is an interesting connecting bnk between t he- 

 other Clavelinidse and the Ascidiidse. It forms colonies by gemmation, and therefore 

 belongs to the Clavelinidse, but in the rest of its characters it more closely approaches 

 Ciona than Clavelina. 



The curious form which Philippi described in 1843, under the name of Rh o pal 'cea, must 

 for the present remain unplaced as to family. It was described as a Simple Ascidian, 4 

 but the account given of it and the figures published certainly suggest that it belongs 

 to the Clavelinidse, and will probably find a place near Ecteinascidia. In opposition to 

 this view, the internal longitudinal bars of Rhopal&a are figured as being distinctly 

 papillated, a condition of things of which I have never seen a trace in any specimen of 

 Ecteinascidia. For a determination of the affinities of this interesting species we must 

 be content to w T ait till we have some information as to the important point of whether it 

 possesses the property of reproducing by gemmation. 



Setting Rhopalaa aside, there remain three genera which undoubtedly belong to this 

 family, viz. : — Clavelina, Ecteinascidia, and Perophora, each of which is represented by 

 several species. The three genera are very distinct, and are easily separated from one 



1 Proc. Roy. Soc. Ediu., 1879-80, p. 714. 



" I have found specimens of Ascidia aspersa, 0. F. Muller, with long stolon-like processes extending from the test 

 at the posterior end and left side of the body. 



3 e.g., frequently in large individuals of Ciona intestinalis, Linnaeus. 



4 This was subsequent to the publication of Milne-Edwards' Memoir creating the group of Social Ascidians. 



