AND HABITS OF THE AMMOCHARID. 247 
an hour or more; and then a swelling of the tube suddenly 
commenced, at a point about a millimetre distant from the 
fractured end. This internal swelling gradually separated the 
sand-grains far apart, and, through the interstices, it could be 
seen that the branchial tentacles were tightly screwed together 
to form a solid living cone, and that the fore part of the 
“thoracic” segment was so fully distended by the coelomic 
fluid, that all trace of a division between it and the base 
of the branchial crown had disappeared. The worm kept up 
a constant revolving motion, the bristle-bundles being actively 
at work, whilst, between the ventral tentacles, both ends of the 
“ Lippen-organ ” (J.0., Pl. 23. figs. 1, 3, & 7) simultaneously rasped 
the membranous tube, their ventral lobes striking downwards, 
and their dorsal lobe vigorously pulling in the opposite direction. 
In a few minutes the tube was in this way burst and cut through, 
the severed part or ring being gradually thrown otf by the 
tentacles, which, quietly unfolding themselves one by one, with- 
drew from the interior and passed the ring over their tips. 
Freed from the stony encumbrance, the branchial tentacles were 
seen to be naked and fully exposed to the water; but careful 
illumination showed that the “thoracic” segment was once more 
invested with a new membranous, tightly-fitting covering, so 
beautifully transparent as to be almost invisible, and this I 
found to be the foundation layer for the new conical tip. It 
had been formed while the animal was sheltering and revolving 
in the old tube; the necessary secretion being doubtless supplied 
by the first two pairs of thread-glands, and passed forward, by 
and over the tentacles, which, as I have stated, were screwed 
together so as to form a conical mould. On one occasion, a 
tentacle was seen to be extended beyond the rest, as though to 
form a pivot to keep the minute aperture at the tip of the 
tube open. Immediately the tentacles were freed from the stony 
ring, the animal proceeded to complete the formation of the 
membranous tip. It advanced so far in its tube that the whole 
of the branchial crown was exposed, and the new tip, as a trans- 
parent skin, was fully stretched over the “ thoracic” segment. 
Then, for the space of five minutes, the first two bundles otf 
bristles worked vigorously, like brushes, spreading the cement 
which, with accompanying violent contortions of the body-wall, 
was poured out by the thread-glands, the animal meanwhile 
keeping up a constant, steady revolution within its tube. Then 
