350 NATURAL HISTORY. 
cula and thrust down into the capacious stomach. So 
voracious are these animals that they will attempt to 
swallow articles which their stomachs can not possibly 
accommodate. In this case the animal will perhaps hold 
the mass partly in and partly out of the stomach firmly 
with its tentacles, pushing it farther in as fast as the low- 
er part of the mass is digested. 
609. In § 599 I spoke of the arrangements of the ten- 
tacles of different animals. The structure of the tenta- 
cles of the Actinize is very peculiar. Their power of 
holding on is owing to a multitude of cells, in which 
there are coiled up in a spiral form fine wire-like fila- 
ments. These can be shot forth from their cells to a con- 
siderable length, and this being done with a multitude 
of them enables the animal to hold on fast to its prey. 
610. Some of the Polypes of this order have a skele- 
ton. It is formed inside of the animal at its lower part, 
and it is fastened to the spot where the Polype lives. 
We may consider it as a foundation frame-work for its 
body. Resting on this, it puts forth its arms continually 
to take its food. 
611. But this skeleton differs from the skeletons of all 
other animals in one respect. Other animals retain their 
skeletons all their lifetime; but the Polype does not. It 
is constantly making new skeleton. It is a singular proc- 
ess, and I will describe it to you with its results. The 
very lowest part of the Polype is continually dying, and 
with it the skeleton which it covers. But as this dies 
the animal keeps its full size, for the body is continually 
supplied with new living substance on the borders of the 
dying portion. It grows just as fast as it dies. It there- 
fore is all the time moving upward, making new skele- 
ton, and leaving the old below. The result, you plainly 
see, would be a column of dead skeleton with the Polype 
at the top of it. In this column, after a while, the living 
part is but small in comparison with the dead part below. 
612. This result you see represented in Fig. 272, one 
