216 HIROSHI OHSHIMA 
sradually widens and at last flattens out, so that the tentacles 
freely protrude above the body-surface. 
Alimentary Canal.—the rudiment of the gut has been 
growing both in length and diameter by rapid cell-division 
and by increase of the cells in height. Its anterior end extends 
beyond the ring canal by which it is encircled (Pl. 9, fig. 25, g), 
and while both ends remain solid its middle portion has a dis- 
tinetly discernible flat lumen lying parallel to the frontal plane 
(Pl. 9, figs. 24 m, F; 9). 
Enterocoeles.—Karly in the dololaria stage, where 
four of the primary tentacles have become apparent, the nght 
and left enterocoeles come into contact with each other at their 
free margins. Both the enterocoeles have been rapidly growing 
in size, extending across the median line and encircling the gut. 
The fusion of their ends takes place on the right side, beginning 
either at the anterior part or at the posterior part of the line 
of contact, leaving for a while an oblique incision at either end 
of the line (PI. 9, figs. 21, 23 ; re, le). 
The other ends of the two vesicles approach each other but 
are separated by a narrow interval. This intervening part gives 
rise to the dorsal mesentery in the end, and lies at first obliquely 
on the left side, beginning anteriorly near the mid-dorsal line to 
end near the mid-ventral line. It, however, gradually bends ito 
an S-shape, indicating the three sections as found in the future 
mesentery—the first, mid-dorsal and descending section; the 
second, oblique and ascending section on the left; and the 
third, descending section running along the mid-ventral 
line. 
IT failed to find any ‘ finger-hke process’ as seen by Bury 
(6, p. 48) in Synaptids and verified by others. Though very 
often there appears a process on the antero-dorsal end of the 
enterocoele, stretching beyond the primary stone-canal to the 
left, I could not follow its fate, and am uncertain whether 
the peripharyngeal sinus originates from it or not. 
The behaviour of the enterocoeles in C. planei was first 
observed by Selenka (45, p. 171), accordmg to whom the 
union of the right and left vesicles takes place on the ventral 
