THE METAMORPHOSIS OF ECHINODERMS. 61 
and then in paraffin—the latter alone, in my experience, gives 
no satisfactory results with any of the pelagic forms of Echino- 
derm larve. 
The “ oral cavity ”’ (as I propose to call the anterior prolon- 
gation of the left body-cavity) closely follows the movements 
of the hydrocel. Its anterior end is thus bent round the 
dorsal surface of the cesophagus, and down on the right side ; 
while its posterior end, where it joins the left body-cavity, is 
brought down by the movements of the hydroccel nearer to the 
ventral surface and to the middle line. This end of it, however, 
does not correspond with the extreme posterior end of the 
hydrocel, but passes on to the surface of this organ at first 
between the polian vesicle and the second tentacle, and after- 
wards, when the polian vesicle has about reached the middle 
line, between the second tentacle and the first radial vessel. 
At this point, towards the end of stage A, it separates com- 
pletely from the left body-cavity. Its two ends then come 
nearer and nearer together, and, as I believe, fuse together 
early in stage B, so as to form a complete ring on the oral 
surface of the water-vascular ring (figs. 11 and 15). It is just 
possible, indeed, that the two ends do not fuse, but that a thin 
mesentery remains separating them; but I have not been able 
to find anything of the kind. 
The subsequent history of this cavity I have not traced. It 
may be identical with the cavity figured by Semon in the same 
position at a later stage (82, pl. iv, fig. 9), but none of my 
specimens are old enough to determine this. 
Besides this oral cavity, we noticed in Auricularia another 
process of the left body-cavity overlapping the anterior end of 
the right body-cavity on the ventral side (fig. 3). This process 
(hardly recognisable in Auricularia) increases rapidly in size 
during stage A, and may be called the “ventral horn” of the 
left body-cavity. It grows rapidly round the right side of the 
cesophagus, on to the dorsal surface (figs. 10 and 11). The 
wall separating it from the right body-cavity very soon breaks 
down, starting from the ventral middle line; but up to the end 
of stage A a small portion of it remains (figs. 11—18), which 
