966 prof, owen ox the [Nov. 19, 



external or internal part of the shell : in some rare cases (Cryptoceras) 

 the siphuncle is ectomarginal, as in Ammonites ; and in still rarer 

 instances (Clymenia) it is entomarginal, as in Spirula. There are 

 also species in which it begins by being marginal, and gradually shifts 

 to the more typical position as the shells grow and the number of 

 the septa increases. 



In the Nautilus imperialis, e. g., the siphuncle is at first, i. e. 

 along the first twenty chambers, entomarginal, or near the concavity 

 of the shell-curve, as in Spirula ; but in opposite relations to the 

 back and belly of the animal. After the twentieth chamber the 

 siphuncle gradually gains the central or excentral position 1 . Nautilus 

 ziczac shows a similar structure. The immature position of the 

 siphon in the existing Nautilus was longer retained in the old 

 Tertiary species. 2 



Various suggestions have been made as to the efficient as well as 

 final cause of the successively vacated parts called " chambers," with 

 their partitions and connecting siphon, in the polythalamous and 

 siphoniferous shells of Cephalopods. 



In elucidation of this question, much mooted by different writers 3 



Fig. 1. 



Vermetus gigas. 

 Section of chambered part of shell. 



after the publication of my Memoir on the Pearly Nautilus, I 

 adduced the instances of such vacuities or chambers in the Mollusca, 



1 Catalogue of Hunterian Cephalopoda, 4to, no. 137, p. 32. 



2 See the just remarks of Barrande on the variation of the siphuncular cha- 

 racter in certain nominal species (Cephalopodes Siluriens de la Bohenie, 8vo, 

 1867, p. 24). 



3 Buckland, ' Geology and Mineralogy considered with reference to Natural 

 Theology.' 1835, vol. i. p. 329. De Blainville, " Sur l'usage du siphon des 

 coquilles polythalames," Annates franchises et etrnngeres d'Anatomie et de 

 Physiologie, t.i. 1837. 



