448 F. A. Bather — Qroicth of Cephalopod Shells. 



stiffened in definite tracts by calcai-eous discs to form an anterior 

 "shagreen field" and a posterior "mucro"; iii. Inner portion, 

 Ycntral ; chiefly consists of the so-called " spongioid tissue " ; in 

 sagittal section the organic membranes forming this are seen to be in 

 groups ; in each group the ventral membranes lie close together and 

 are supported by a calcareous lamina of prismatic structure, thus 

 forming a " lamella " ; the remaining membranes are stretched freely 

 at regular intervals ; between the lamellee, at right angles to them, 

 are calcareous " sinuous partitions," which radiate towards the 

 margin ; through these the free membranes pass, and are by them 

 supported : decalcification shows that all the membranes are con- 

 tinuous with those of the Middle plate (Fig. 2), 



(yS) The Outer portion is secreted by the mantle ; its form 

 is determined by that of the Inner portion. Secretion does not 

 explain the growth of the Inner portion : the last formed lamellse 

 lie more closely on one another than do those of earlier age ; 

 the spaces between the freely-stretched membranes are not 

 smaller, but the number of those membranes is less, they are 

 absent from at least the hinder edge of the last formed lamella ; 

 hence the partitions are shorter, and the space between the insertion- 

 lines of the two last lamellae is less than in other parts of the shell 

 (Fig. 2) ; each partition is bifid ventrally, this end is the first to 

 appear. It follows that the membranes are separated from the 

 lamellee, not secreted by the mantle ; and the calcium carbonate is 

 only deposited on the inner, i.e. dorsal, side of the conchyolin 

 (Fig. 3) : the subsequent sundering of the lamellae is accounted for 

 by intussusceptional growth of the shell. 



(fy) The Outer portion equals the Belemnite guard ; the Inner 

 portion the phragmocone, the lamellse being septa ; the Middle plate 

 was represented by an epicuticula of the phragmocone : this is con- 

 firmed by fresh details. Hence in all chambered shells the hinder 

 end of the body is in continuous contact with the youngest septum, 

 and the subsequent sundering of the septa is effected by the growth, 

 of those zones of the shell- wall that lie between the sutures. 



Thus far Riefstahl. 



Undeniably there are objections to the Secretion-hypothesis, 

 such as the complex structure of nacre, especially when freely- 

 stretched membranes are present. But neither alternate secretions 

 of lime and gas or, as in Sepia, liquid, nor alternate deposition of 

 lamellse and partitions are so absurd as Riefstahl maintains. Despite 

 their regularity, due to advance of structure and function, we need 

 not think the septa of Cephalopods originally different from those of 

 some Gasteropods. Still the chambers of a Nautilus have, thanks to 

 the siphuncle, a vitality denied to the septate portion of a Bulimus ; 

 here Eiefstahl's conclusions harmonize with those of F. Miiller on 

 shell formation in Lamellibranchs : " The shell is no product of 

 secretion, but has independent life, and grows by intussusception " 

 (Zool. Beitr. Breslau, 1885, p. 206). Yet the avascular nature of the 

 shell renders it hard to credit elaboration of lime in parts removed 

 from the animal, nor is it to be thought that the latter is wholly idle ; 



