MOLLTJSCA — CEPHALOPODA. 
147 
active predatory animals, or wliether they were sedentaiy 
and possibly attached by the shell, is uncertain. l>nt ceitain 
it is that at an early period the hump was drawn out into a 
lon<T visceral cone, and that the shell ac(]^uired a similar 
shape. Then followed a mode of growth very common in 
sedentary animals that form a tubular shell, and already 
observed in corals, worms, bryozoans, and gastropods. The 
mollusc continued to build up the shell around its opening, 
and thus formed a long tube. As the animal moved along 
this tube, the visceral cone was pulled away from the shell- 
Fig. 79. — Primitive Cephalopod shells: Endoccras. a, The end of a shell, 
broken at its apex ; b, the same cut in half showing the chambers (sc), 
the swollen end of the wide neck-tube (cc), and calcareous substance 
deposited by the end of the siphuncle (cd). c. Fragment of another 
shell cut in half, showing the chambers separated by septa (.s), the 
largo neck-tube (sc), a sheath (sh), and the endosiphou (cn). Natural 
size, (a, b, after Holm ; c, after Foord.) 
wall ; but its skin went on secreting shell-substance, and 
formed a partition shutting off the space between the visceral 
cone and the outer wall. In Endoceras, here shown (Fig. 79), 
the visceral cone remained attached to the end of the outer 
shell, and shrank at a little distance from the apex, so that 
the partition or septum does not go right across the shell, but 
shuts off a chamber at the side. Shrinkage then took place 
a little higher up, and another chamber was formed. By 
the continuation of this process there arose a series of 
chambers entirely shut off by septa, and the down-turned 
portions or necks of these septa formed a long tube in which 
Gallery 
VII. 
