146 GUIDE TO THE FOSSIL INVEETEBEATE ANIMALS. 
Gallery 
VII. 
Table-case 
1 . 
(Fig. 78 (1) the ann.s are re}treseuled by short lobes, and those 
are iiirnished with tentacles that can l)e Avithdrawn into 
sheaths. The eyes in tlic long-armed forms are conspicuous 
and are as liighly develojjed as tho.se of a vertebrate animal ; 
in the nautilus one can detect only a small o]>ening leading 
into a chamber that acts like the simple optical apparatus 
known as a ])in-hole camera. If the mantle-cavity be opened 
and the gills exposed, it will be seen that there are four gill- 
plumes in the nautilus, hut only two in the other forms 
mentioned, and indeed in all other cephalopods now living ; 
for this reason, and in view of the other diflerences. Sir 1>. 
Owen divided the Cephalopoda into two divisions : Tetra- 
branchia (four gills) and ] )ibranchia (two gills). We know 
that some fossil cephalopods, clearly related to the modern 
1 )ibranchia, had long arms, suckers, and large eyes ; hut we 
cannot be certain that all of them had only two gill-plumes. 
Other fossil shells resemble that of the nautilus, and in some 
of these cases the animal most probably had four gill- 
plumes ; hut in other cases there are no grounds for any 
such sissertion. Therefore Owen’s classification is unsuitable 
for fossil cephalopods. 
In order to arrange systematically the large number of 
extinct cephalopods one must consider chiefly those parts of 
the animal that can be preserved in the rocks. The beaks, 
the horny rings of the suckers, and similar structures have 
already been mentioned (p. 123), as also the fact that some 
fossils found in very fine clays have preserved even the 
muscles of the mantle (Fig. 86 a). As a rule, however, one 
finds only the shell, with which most of the Cephalopoda are 
provided. This shell has in the course of geological time 
undergone many changes, and has been modified in several 
directions. It is moreover intimately related to the structure 
and development of the whole animal, and therefore furnishes 
an excellent basis of classification. 
Specimens, models, and drawings have been arranged to 
show the history and relations of the cephalopod shell, 
and to these attention should first be directed. As in all 
Mollusca the shell is primitively secreted by the mantle or 
skin of the visceral hump, and, at its edges, by the back- 
wardly turned folds of the mantle. Originally then the 
.shell follows the shape of the visceral hump, and we may 
suppose that, in cephalopods older than any which are 
known to us, it was a somewhat conical cap, not unlike the 
shell of some uncoiled gastropods. Whether these were 
