On the Origin of the Ascidian Mouth. 



455 



Willey (1893) describes the early closure of the neuropore and the 

 formation of the mouth by the fusion of the stomodseal invagination with 

 the pharynx. Shortly (in Clavelina) or some time (in G'iona) after the forma- 

 tion of the mouth, the hypophysial canal connects up with the base of the 

 stomodreum. He does not appear to be perfectly certain of the way in which 

 this latter connection is made. 



Seeliger (1904) considers that the neuropore closes early and that the 

 neural tube separates from the ectoderm. Later, the hypophysial canal 

 separates from the sensory vesicle, and grows forward to unite with either 

 the stomodaeum, the pharynx, or the junction of the two. But he does not 

 seem to think that the details have been satisfactorily established. 



As Willey considered that the point where the hypophysial canal broke 

 through into the stomodseum corresponded with the previously closed neuro- 

 pore, my conclusions are somewhat in harmony with his. 1 have not had 

 the opportunity of studying the origin of the mouth in Ciona or in a 

 Phallusiid, and possibly in these the neural tube separates from the ectoderm. 

 But in Clavelina, which was studied by Willey and Seeliger, I do not believe 

 that the separation occurs. Willey did not consider this point. Seeliger 

 stated that separation occurred, but I am inclined to think that he did not 

 actually investigate it in Clavelina. He states, however, that he examined 

 larva? in which the opening of the hypophysial canal was present, but in 

 which the mouth was not yet broken through. 



I have been unable to find a stage in which the canal is connected with 

 the pharynx and not with the exterior. If an actual lumen be implied, 

 Seeliger's statements are confirmed by my findings. But the essential thing 

 is the arrangement of the cells to form a tube, whether closed or open. 



When observations are made on entire embryos, the connection of the 

 neural tube with the ectoderm is easily overlooked. Also the fixation of 

 the material or the subsequent handling might break the connection. My 

 material was fixed with Zenker's fluid, and the embryos remained during 

 preservation, embedding and cutting in situ inside the adult. This treatment 

 prevented any separation of the parts of the embryo, but made the examination 

 of the sections somewhat difficult. 



In the early stages of development the neural tube is widely open in front, 

 but at the time when the tail grows out, the neural tube is closing. The 

 margins of the neuropore come together and from the outside little if any 

 indication remains of the position of the neuropore. In Clavelina there is 

 usually a slight depression, more or less filled by testa cells. A section of 

 this stage (fig. 1a) shows the neural tube still connected with the ectoderm, 

 but a definite lumen cannot be traced to the surface of the ectoderm. The 



