460 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



During the last of the embryonal period and the free-swimming larval exist- 

 ence the parietal canal undergoes important changes. It soon closes off entirely 

 from the hydroccele and becomes a tubular structure running dorsally from it, 

 while in the free-swimming larva its major part comes to lie directly in front of it. 

 The left end grows posteriorly and lies at the left of the hydroccele and gut, 

 reaching the ectodermal body wall just before the fourth ciliated band, eventually 

 opening outward here through the pore. 



In the young larva the pore lies between the two branches of the hydroccele, 

 between the places where the first and fifth evaginations will arise. 



The formation of the pore often occurs in the final embryonal period, and the 

 opening is always clearly visible in larvae of a few hours' age. There is some- 

 times a slight infolding of the body wall in the immediate vicinity of the pore, 

 but again the small-celled terminal portion of the parietal canal may extend into 

 the body wall as if it had bored its way through it. In any case, the part taken 

 by the ectoderm in the formation of the end piece at the pore is very small. 



At its first appearance the pore lies to the left of the median plane, distant 

 from it about one-sixth of the circumference of the larva. Later the distance 

 increases somewhat. 



At the pore the lumen of the canal is very small, but it broadens rapidly 

 inward and anteriorly, going over into the transversely running major portion, 

 which is very differently developed in different individuals. In somewhat more 

 developed stages in which the anterior gut end is more rounded it becomes 

 more extensive, and at the same time part of the earlier ascending branch retains 

 a more transverse course. 



At the place where these two sectors pass over into each other there is to be 

 noticed in quite young larvae a short hornlike process, directed toward the right 

 and posteriorly, which lies in the angle between the anterior end of the gut and the 

 left hydrocffile. This process, which later disappears, appears to be the last rem- 

 nant of the original connection between the parietal canal and the hydrocoele sack. 



The last section of the parietal canal is the anterior process. This arises from 

 the transverse portion ventrally near the median plane and runs forward, a little 

 toward the left, extending to before the second ciliated ring under the attachment 

 pit. Anteriorly the lumen becomes progressively narrower, finally disappearing 

 entirely, though in some cases the foremost end swells into a knoblike structure 

 with a prominent lumen. 



The posterior end, representing the position of the transverse section, lies 

 very near the base of the vestibular invagination, though not directly upon it. 

 Farther forward the tube approaches it and clings to it, so that the whole process 

 is feebly curved in the form of a bow with the concave side directed ventrally. 

 This anterior process is a very striking structure. 



The wall of the parietal canal is composed of a single layer of cells through- 

 out; it is markedly thinner than in the 75-hour embryo (the stage in which it 

 was last described), though everywhere clearly demonstrable and sharply differ- 

 entiated from the other tissues. The cells are cubical or slightly flattened, some- 



