Page:EB1911 - Volume 05.djvu/721

 mantle being enclosed by the shell, the shell is enclosed by the mantle. The earliest stage of this process is shown in the recent Spirula, though it is perhaps not impossible that in some of the later fossil Ammonoids the shell was becoming more and more internal. The shell of Spirula (fig. 18) is coiled somewhat like that of Nautilus, but the coils are not in contact, the direction of the coil is endogastric or ventral instead of exogastric, and the shell is very much smaller than the body. Like that of Nautilus it is divided by septa and traversed by a siphuncle. The relation of the animal to the terminal chamber is as in Nautilus, but the body extends far beyond the aperture, and folds of the mantle grow up over the shell and cover it everywhere except part of the dorsal and ventral surfaces.

The next modification in the enclosed shell is the addition to it of secondary deposits of calcareous matter, by the inner surface of the shell-sac. Successive layers are deposited on the posterior part of the original shell, whether coiled or straight, and these layers form a conical mass, which may attain great thickness. A somewhat coiled shell with such a deposit is seen in Spirulirostra (fig. 17, C) of the Miocene. In the next stage of modification secondary secretion forms a long and broad projection of the dorsal lip of the aperture; this is well developed in the belemnites (fig. 19). Thus in these modified shells three parts are to be distinguished: the original septate shell, which has been called the phragmacone; the posterior conical deposit, called the rostrum or guard; and the anterior somewhat flat projection, called the proostracum. In the living Dibranchiata other than Spirula the phragmacone and rostrum have become very rudimentary. The shell of Sepia (fig. 20) consists almost entirely of the proostracum, the little ventral hollow posteriorly representing the phragmacone, and the posterior pointed projection, the rostrum. In the Oigopsida the shell is represented by a proostracum which is no longer calcified by forms a chitinous plume or gladuius, and a similar rudiment occurs in Loliginidac (fig. 21) and Sepiolidae. Lastly, in the Octopoda the shell is represented only by small chitinous rudiments to which the retractor muscles of the head and funnel are attached; these are paired in Octopus, unpaired in other cases as in Cirrhoteuthis.

The early appearance of the sac of the mantle in which the shell is enclosed has led to an erroneous identification of this sac with the primitive shell-sac or shell-gland of the Molluscan embryo. The first appearance of the shell-sac in Dibranchiata is shown in figs. 35, 36. Its formation as an open upgrowth of the centro-dorsal area, and the fact that it appears and disappears without closing in Argonauta and Octopus, was demonstrated by E. Ray Lankester.

In Argonauta (the paper nautilus) the female only possesses a shell, in which the body is contained; but this is not