ox THE MICROSCOPIC STRUCTURE OF SHELLS. 113 
41. But at one revolution backwards they were — outer 5, central 5, inner 
19; total 29.” This fact harmonizes well with the preceding explanation ; 
for it simply indicates that whilst the tliickness of the whole shell undergoes 
a considerable increase near the lip, that increase is due to the greater de- 
velopraetit of the outer and middle layers, the inner layer being positively 
thinner than it is at a distance from the lip, where it has been thickened by 
aucoessive deposits. I can see no reason to attribute the foregoing diiferences 
in the relative thickness of the three layers at different parts of the shell, to 
any absorption or removal of the outer layers, as Mr. Bowerbank seems in¬ 
clined to do: for it seems impossible to Imagine tlint the external and 
middle layers can be tbinned by absorption, witliout the removal of the inner 
layer; since any absorbent action must lake place from within, being effected 
by the surface of tlic maiitlo. I am far from denying that such absorption 
do«s take place ; but the explanation is not applicable to the facU just cited, 
which seem to me to indicate clearly that the formation of the inner layer is 
progressive and not completed at one efi'orL The truth appears to me to be, 
that whenever an addition is made to the siicll, the outer and middle layers 
are simply joined-on to the edges of the old margin, but that the internal 
layer is carried backwards foracertain tlistance into the interior of the shell, 
where the new formation forms a lining to the old, and ioert-ases its thick¬ 
ness, just as in the Bivalves. I have never been able, however, to trace it 
very tiir back; and it certainly can seldom or never line the whole shell, as 
Jt does ill most Bivalves, But this new layer seems to cover that part of the 
internal surface which is in contact with the moving parts of the animal; 
and thus serves to prevent that irregularity which could scarcely fail to 
«ist, where the new internal layer, Tike the middle and external, simply 
joinod-ou to the edge of the preceding. 
94. In Mr. Gray's paper already referred to, a number of facts are 
brought together, which prove beyond all doubt that absorption of shell- 
suhstance may take place; but this, like the abrasion of stony matter by the 
bwiiig MolluacH, is a surface-action merely; and the complete absence of 
vascular or tubular communication between the mantle and the remoter 
layers quite furbids the idea that any such process can be effected without 
the removal, in the first instance, of the layer in immodiato contact with the 
oianUe. This removal takes place to a great extent in some of the spiral 
t^uvalvcs; especially in such as Conus, in which the older portion of the 
. entirely wrapped round by the newer or outer whorls. In such cases 
1* (as remarked by Mr. Bowerbank) on the otiUr layer of the involved 
the absorbing action is exerted; tins layer being removed, the 
w attacked; and even a portion of the inner one is often thus 
? f^vbed, that which remains (in casea where but little strength is rerjuired) 
much thicker than writing-paper, quilt; translucent, and exhibiting 
” ^L- Jirecisely the same aiipearancc as that presented by a 
section of this layer, taken parallel to its surface. 
IBM * now pass in review some of the more interesting cm^es of de- 
iho i 1 regular type of conformation, and shall commence with 
I/dourancluata. In Patella the inner and outer layers arc of the nor- 
• ^‘'■“riure, though rather leas compact than usual; the lamin® of the 
cr layer especially Iwing very large -and irregular, and by no means firmly 
querent to one another. The middle layer, however, has a very different 
onW composed of tolerably regular polygonal cells, which form 
hn ^ tI' parts, whilst it» otliers they arc elongated into 
bv ^ayer is shown at fig. 51 a, as it is brought into view 
. ^ction of the wide part of the cone nearly parallel to the surface. The 
