312 . NEW YORK STATE MUSEUM 



also connected at its convex side to the nearest wall by a band of 

 crystals of organic carbonate of lime. The interspaces are not only 

 arranged symmetrically, but also delimited so sharply by uninter- 

 rupted lines, that it is hardly to be doubted that the calcite bands 

 connecting the endosiphocone and wall of siphuncle are the remains 

 of the membranes which held the visceral cone in position within the 

 siphuncle and probably became partially calcified during the lifetime 

 of the animal. The interspaces remained cavities till they were 

 filled by the large calcite crystals now occupying the siphuncle.^ 

 The supposition of the fixation of the visceral cone and inclos- 

 ing endosiphosheath to the ectosiphuncle, finds support in the 

 occasional presence of bands of gray brown limestone, extending 

 from the endosiphocoleon (virtually the continuation of the vis- 

 ceral cone) or more posteriorly from the endosiphotube, to the wall 

 of the siphuncle. Such a section is reproduced in plate 8, figure 5. 

 The horizontal transverse band with the inclosed endosiphotube is 

 evidently the '' endosiphoblade " of Holm. This is held in a manner 

 corresponding to the fixation of the endosiphocone described above 

 by a band that is placed perpendicular to the endosiphocoleon.^ The 

 extension of the internal space of the visceral sac (endosiphocone) 



^In this particular siphuncle the interior is 20 mm from the end of the 

 endosiphocone already so calcified, apparently by secondary calcification, 

 that hardly any trace of the endosiphocoleon is left [see pl.8, fig.8]. 



'These supporting membranes were, as we have mentioned above, 

 recognized by Dewitz and more fully described by Holm. The latter 

 author [/. c, 1887, p. 16] sums up his observations on these supporting 

 membranes in Endoceras gladius in the following statement: 

 " During the retrogression of the siphon in the siphuncular tube there were 

 secreted by the siphon three longitudinal membranes which were prob- 

 ably soft, pliable and extended to the wall of the siphuncular tube, one 

 from each of the angular marginal edges and one from the median line of 

 the convex side. Their function was probably to fix the end of the 

 siphon, which was suspended in the siphuncular tube in a position in the 

 middle of the latter. A similar organ was, as we have seen above, ob- 

 served by Dewitz in the siphuncular tube of a specimen of "Endo- 

 ceras commune." In consequence of this structure the " Spiess " 

 maintains in all specimens of the species in question, which have been 

 investigated by me, the same position in the middle of the siphuncular 

 tube and indicates an invariable position of the end of the siphon. The 

 thin (cuticular) membranes were secreted along the whole length of the 

 siphon." 



