AND STRUCTURE OF THE BELEMNITE. 409 



some of the calcareous specimens ; for when the cone is bro- 

 ken across, it separates by these lines, leaving in the upper part 

 a concavity, and on the lower a corresponding convexity, and 

 exhibiting a spot occupied by a siphunculus or duct, like the 

 nautilus and other shells, at the side of the concamerated cone 

 which is nearest to that of the belemnite ; the use of which, 

 according to Parkinson* vol. iii. p. 129. " in all the multilocu- 

 lar shells, was to bring the animal to which they are appended, 

 with its shell, to a degree of specific gravity, so near that of 

 water, as to render it capable of being raised or sunk with faci- 

 lity, by the apparatus of its siphuncle." I conclude that this 

 part of the belemnite must have existed in such a state, as, by 

 its lightness, it must, like the closed chambers, have served as 

 a float to the animal. 



Whether this was the use for which the organization of the 

 belemnite was destined, entirely depends upon the nature of the 

 animal to which it was appended. Any conjectures on that head 

 must therefore be wholly gratuitous. One thing is evident, 

 however, and marks a conspicuous distinction between the 

 concameration, if I may use that expression, of the belemnite 

 and the nautilus : the chambers of the last are separated from 

 each other by a firm compact shell, while nothing of the kind 

 appears to have existed in the belemnite. Its organisation may 

 have been composed of a soft membranous substance, easilv 

 removed on the animal being deprived of life. 



This supposition is somewhat confirmed by the appearance 

 of the belemnites found in the Antrim chalk, which must 

 have been dead shells (as that expression is understood by na- 

 turalists) a." the time they were inclosed in the strata ; for, be- 

 sides all appearances of the body to which they had belonged 

 being totally obliterated, the fossil itself is not only fractured 

 at the edge of the alveolus, but is frequently, though not al- 

 ways, found to have been perforated by serpulse, and thus 



3 f 2 affords 



