in Cephalopoda [Siphonopoda) . 307 



derived from sea-water, is deposited as arragouite upon and 

 between these membranes soon after their secretion ; bj this 

 more purely physical process nacre is produced. The struc- 

 ture of this nacre is like that of the Lamellibranch shell in all 

 essential points (6) ; and if, pace F. Miiller, the present 

 explanation be extended to the shells of all Mollusca, we shall 

 merely be returning to the views of Dr. W. B. Carpenter 

 (1 and 3), with this difference — the membranes are not cel- 

 lular and do not of themselves retain vitality. The membranes 

 of the outer portion are secreted by the mantle-lobes, and in 

 them is formed the Sheath (cap of Orth. truncatum, Barr. ; 

 guard of Aulacoceras and Belemnifes -^ mucro and shagreen- 

 layers of Sepia ; disc of attachment and ligament in Spirala ; 

 outer portion in pen of Teuthidas and gladius of Loligo). 

 These membranes may be rendered entirely or partially 

 spathose by the deposition of lime in them as calcite. These 

 latter facts have been disputed by no recent worker. 



The foregoing explanation accounts very simply for ob- 

 served structures in recent forms if in applying it we remem- 

 ber two principles : — (i.) that secretion of chltin may persist 

 after advance of the body in the shell has slackened ; (ii.) 

 that extent of calcification varies inversely as rate of secretion. 

 Where secretion of chitin is slow in proportion to the growth 

 in length the septa are well separated from one another, and 

 each is calcified soon after the membranes forming it have 

 been deposited ; there is no subsequent growth or movement 

 of the septum (e.g. Nautilus), When, in accelerated deve- 

 lopment, the septa have come close together, calcification 

 cannot keep pace with secretion, and the numerous overlying 

 membranes prevent further deposition of lime in tliose sub- 

 jacent (e. g. Sepia). Remember that calcification is always a 

 slow process compared with secretion, and cannot, like it, be 

 hurried. The growth of the sheath being correlated with 

 that of the shell, we shall expect to find that, in forms where 

 the anterior extension of the shell takes place slowly, the 

 sheath-membranes remain calcified (e. g. Belemaites) , but 

 that in forms where the shell-wall is rapidly prolonged the 

 sheath-membranes cover one another too quickly to permit of 

 calcification (e.g. Coccoteuthis) . Remember that these mem- 

 branes shrink into a very thin space when not kept apart by 

 lime (cf. Belemnoteuthis) . 



Examining extinct forms by the light of these principles, it 

 is easy to see how the various differentiations arose. Nauti- 

 loidea and Ammonoidea considered, let us study the Coleoidea, 

 At the beginning of the Mesozoic Era these split into two 

 branches. In one branch the anterior extension of the dorsal 



