MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 169 



We have still a portion of the coelom to consider — that portion which stretches out in the 

 siphuncle. Upon the right side of the attachment of the genital ligament to the body wall is an 

 opening which leads into a cavity continued to the end of the siphuncle. The structure of the 

 siphuncle has been described in some detail by Haller, and its consideration seems to occur 

 most logically at this point. 



The genital ligament lies directly over the opening from the coelom into the siphuncle, and 

 makes the opening quite difficult to find sometimes. However, it is always single and upon the 

 right side of the ligament. The cavity of the siphuncle is divided into three tubular portions, 

 one of which is dorsal and two ventral. The cavities are lined by a continuation of the coelomic 

 epithelium, which is here composed of low cubical cells. The siphuncle is narrowed where it 

 passes through each septum. At these points the siphuncular cavities also are narrowed, but not 

 occluded. 



The common septal artery passes along the genital ligament to the posterior portion of the 

 body wall near the base of the siphuncle. It here divides into a right and a left branch, from 

 one of which, sometimes the right and sometimes the left, the siphuncular artery is given off. 

 (Text-tig. 11. p. 186.) This extends through the entire length of the siphuncle, finally ending 

 openly. 



The spaces intervening between the cavities of the siphuncle and the siphuncular artery are 

 tilled by a loose reticular connective tissue, the spaces of which contain venous blood, and are in 

 communication through the walls of the body and the various ligaments crossing the c<elom with 

 the cavity of the hsemocoel. 



Haller describes a very curious structure of the epithelium covering the siphuncle exter- 

 nally. According to his description, the basement membrane is thrown into tine longitudinal 

 folds. The epithelium does not cover the edges of the folds, these being in direct contact with 

 the inner wall of the siphon. The epithelium covering the grooves between the edges of the 

 folds forms a continuous protoplasmic layer, in which cell boundaries are not distinguishable. 

 This layer shows striations perpendicular to its surface. It stains intensely with hematoxylin. 

 The widely separated nuclei are disk-shaped and lie in the upper or outer portion of the layer. 



The tissues of the base of the siphon seem always to be continued into the genital ligament, 

 forming a small rounded nodular eminence close to its attachment. In one case, shown in Fig. 38, 

 a cord of tissue extends from this eminence along the ligament of the stomach. What this is I 

 have not yet been able to determine. 



There have been several theories advanced as to the function of the siphuncle. Most con- 

 sider that it is in some way connected with hydrostatic properties of the shell and attempt to 

 explain the rising and sinking of the animal as in some way dependent upon the action of the 

 siphuncle. Reeve's theory of its action is very interesting, and a paragraph from him is worth 

 quoting. 



"The following appears to us to be the manner in which the Nautilus constructs its shell. 

 The animal in its embryo formation deposits a simple hollow shell, out of which it necessarily 

 advances as it increases in bulk; and in order to assist its specific gravity at the bottom of the 

 ocean the vacated portion of the shell is chambered in by the secretion of transverse septa, the 

 animal having first taken the precaution to secure a strong tubiform membrane to the inner wall 

 in order to adjust its position (a consideration of the habits of this pelagic mollusk will show the 

 necessity for this membrane). As the soft parts increase in bulk, the muscular girdle which 

 binds them to the shell would naturally be forced from any adhesion, but from its being fur- 

 nished with a certain degree of elasticity it advances by a series of periodical slips, the sudden- 

 ness of which is undoubtedly counteracted by the attachment of the central membrane. The 

 growth of the shell then proceeds in a circular direction, and serves to buoy up its inhabitant in 

 the water by having the vacated portion chambered in to meet its specific gravity. The geo- 

 metrical increase of it arises simply thus: The natural position of the Nautilus, like other 

 cephalopods, is with its head downward, the shell being consequently above: and the periodical 

 slip of the belt of adhesion most probably takes place when the animal is in this supine position. 

 It lets itself down, and round and round, as it were, upon its axis by the limited extension of 



