Notes on the morphology of the Tunicata. 507 



origin of the rapheal duct, is a region where the basement membrane 

 of the glandular epithelium is absent and the cells of the gland and 

 ganglion merge into one another, being indistinguishable (Fig. 10). 

 This I believe to be a reminiscence of the mode of origin of the 

 ganglion by the proliferation of cells from the dorsal wall of the tube 

 in the larva, which is at first the neural tube, but which becomes 

 later the definitive duct of the neural gland (cf. Willey, 1893, II). 

 Similar relations are found in Rhopalaea neapolitana, in PeropJiora 

 viridis and in many other forms to some of which reference will be 

 made later. 



In addition to this fusion of gland and ganglion, we observe 

 (Fig. 10) on each side of the duct of the gland that the peri- 

 ganglionic membrane is wanting and that certain of the large cells of 

 the ganglion push down on each side of the duct. Further forward 

 these ventral outgrowths from the ganglion are much larger and 

 push down around the duct completely surrounding it, and are so 

 closely united to the gland as to present no line of demarcation from 

 the latter.. As seen in Fig. 10 these downgrowths are partially sepa- 

 rated from the ganglion by an intervening membrane. 



Similar ganglion cells lying outside the contour of the brain and 

 often separated by a membrane from the brain , are found in most 

 Ascidians and in Salpa above the gland. I have not found sufficient 

 indication that these ganglionic outgrowths innervate the gland. 



Dia^ona violacea Say. 

 Plate 34, Fig. 11. 

 Diazona violacea shows conditions so similar to Clavelina rissoana 

 that no description is needed. I should, however, say that in this 

 species I have been unable to find any fusion of gland and ganglion, 

 and also that there is no such great development of the ganglionic 

 downgrowths from the brain. There are merely the usual few large 

 ganglion cells on each side between the gland and the brain. 



Ecteinascidia turbinata Herdm. 

 The general relations in this species are so similar to those in 

 Clavelina that they need little description. One noteworthy difference 

 is the fact that there is no rapheal prolongation of the duct of the 

 neural gland. The rapheal nerve is present and is richly gangliated. 

 It arises from the posterior end of the ganglion just above the blind 

 posterior end of the duct of the gland. 



