56 BRITISH TUNICATA. 



in it, ai a very early period, the cloaca freely communicates 

 with the water-space or atrium on each side. At the same 

 time it must be allowed that it is more than probable that, at 

 the earliest stage of existence, in Clarelina and in other Tuni- 

 cates the great ventral channel is united throughout to the 

 wall of the pallial chamber; and hence the statement of 

 Krohn does not seem at all unlikely. And, moreover, we 

 thus learn that this great blood-channel is developed in con- 

 nexion with the lining membrane, with which it continues 

 ever afterwards more or less connected. We have already 

 seen that the transverse or primary vessels take their origin 

 in this same vessel, and that they in their turn give oft' the 

 secondary vessels : these are the essential parts of the bran- 

 chial tissue, and when we look to its anatomical structure as 

 well as to its mode of growth, we can scarcely doubt that the 

 network of the gill is truly vascular. Speaking, therefore, of 

 the branchial sac as a perforated membrane, as is frequently 

 done, gives an erroneous idea of its apparently true nature. 



The longitudinal bars which have been so frequently 

 alluded to, and which lie in a plane a little above the inner 

 surface of the respiratory sac, are non-essential parts of the 

 organ, their function apparently being, as previously stated, 

 to protect the surface of the gill, and, by the aid of their 

 cilia, to sweep the alimentary matters towards the oral 

 lamina. They are not always developed ; they do not exist 

 in Clctveliiia- neither are they apparently present in P<-n>- 

 phora and they seem to be absent in several of the com- 

 pound Ascidians; in DnUolum they have likewise disappeared. 



It is stated above, that ChiveUna is nearly related to Salpa; 

 but Pyroso-iita and Doliolum come much nearer to it in their 

 general structure, as well as in the details of their organisa- 

 tion. Unfortunately I have never seen either of these two 

 interesting forms but, judging from the able descriptions of 

 them by Prof. Huxley in the ' Philosophical Transactions/ 

 they both present examples of imperfectly-developed gills. 

 In Pyrosoma the secondary vessels are entirely absent, and 

 the primary vessels of the two lateral laminas of the branchial 

 sac do not reach the endostyle, their development having been 

 arrested before they extended so far across the respiratory 

 cavity ; their distal extremities, however, will undoubtedly 

 open into the system of pallial sinuses ; in no other way can 

 the flow of the blood through the gill be explained : the cir- 

 culation is therefore to this extent embryonic. " The longi- 

 tudinal bars " of Huxley are the homologues of what have 

 been so designated throughout this communication, and are 



