106 HENRY BURY. 
Turning to fig. 48, we see that a gradual rotation of the 
water-vascular ring, as it moved into the left side, would 
suffice to maintain the straight course of the water-tube, 
without any change of interradius of either of its extremities 
—the total angle of rotation being about 70°. Now compare 
Ludwig’s diagram (15, p. 157, fig. v) of the changes by which 
the “ambulacralen Armanlagen” (hydrocel pouches) 
are brought into connection with the “antiambulacralen 
Armanlagen” (formed over the left body-cavity, parallel 
with the mesentery), and it will at once be evident that we 
have here a twisting of the hydrocel through a considerable 
angle (though not as much as 70°) in exactly the same direc- 
tion (remembering that Ludwig gives a dorsal view and I a 
ventral) as I have postulated ; the movement, in fact, may in 
both cases be described as tending to unscrew the water- 
vascular ring from the stomach. 
Here, then, we have a possible explanation of a very remark- 
able phenomenon. Of course so long as we are ignorant of the 
meaning of the extraordinary variability in the point of closure 
of the water-vascular ring noticeable in Echinoderm larve, we 
cannot hope to explain all the peculiarities of development pre- 
sented by Brachiolaria ; but the meaning just suggested for one 
of the most striking of them fits in so completely with the needs 
of my hypothesis that I cannot help attaching a good deal of 
importance to it. In considering why this change is more 
strikingly presented by Brachiolaria than by any other larva, 
we must remember that this form is peculiar in that the 
hydroccel early arrives at its position alongside the stomach, 
while the cesophagus only joins it there at the time of meta- 
morphosis. In Bipinnaria and in Echinids the hydroceel very 
early assumes its final position, and a new cesophagus is 
formed; so that the rotation of the hydroccel (if it is not 
entirely omitted in ontogeny) probably takes place before the 
development of the tentacles enables us to recognise it. In 
Ophiurids the rotation ought to take place exactly as in the 
ancestor; and I think when we come to consider the matter 
carefully we are bound to admit that it does so, though the 
