THE CEYLON PEARL-~OYSTER. 307 
due to simple secretion, arises by a gradual transformation into 
shell-substance of the outer regions of the specialised epidermal 
cells which underlie it, 7, e. in the same manner as the Crustacean 
carapace and the outermost layer of the periostracum. It shows 
a well-marked striation perpendicular to the surface, the striz 
no doubt corresponding to the outlines of the spaces which were 
occupied by the needle-like fibrocrystalline bodies of calcium 
carbonate, and also at times indistinct lines parallel to the surface 
(Pl. XXXIV. fig. 5). In sections in the plane parallel to the 
surface this substance shows an alveolar structure. It sometimes 
shows a tendency to break up into segments, corresponding to 
the underlying epithelium-cells (Pl. XXXIV. fig. 5, hy.’). 
(v.) Hinge-Ligument. 
I do not propose to discuss the hinge-ligament here, as it has 
not the same direct bearing on the question of pearl-formation 
as the above layers, though leathery pearls, composed of this 
substance, are sometimes found (e. g. in JZ. maxima in Australia). 
(10) Tae SHELL-secretinG Eprruecia. 
The ordinary shell-secreting epidermis of J/. vulgaris, so far 
as I have been able to study it in the unsatisfactory material 
available, consists of columnar or tesselated cells (Pl. XXXY. 
firs 8 eps; te 29. 0;ep; 5. Pin XX RV I. firs D1y.06p.5cPl.,, SLM. 
fig. 33, ep.), brick-shaped or palisade- like, according to the 
degree of contraction, in sections perpendicular to the surface, 
with a certain and variable number of goblet- and gland-cells. 
The nuclei of the epidermal cells are oval or spindle-shaped. 
These cells are attached to the subjacent tissues by a basement- 
membrane of delicate fibrille which distinctly marks the 
boundary between the epidermal epithelium and the subjacent 
tissues. Beneath this epidermis is a characteristic granular 
parenchyma (Pl. XX XV. figs. 8&9; Pl. XXXVI. figs. 10 & 11; 
Pi ee eV AL fig, 145, Pl. ane fig. 33, par.), which contains a 
great variety of elements, some being comparable to the 
‘* Rundzellen” and ‘ Langer’schen Blasen” described by List 
(27 b) for the Mytilide, some being dark-staining, apparently 
glandular elements that open out between the epithelial cells 
(fig. 33, gl.). As observed by List (276), this epithelium and the 
underlying tissues are excessively variable in their characters. 
Over the surface of the muscle-attachment the epidermis is 
different (Pl. XX XIII. figs. 4, 4a@; Pl. XXXV. fig. 8, m-ep.). 
Here it consists of columnar cells, usually about 10-12 p» long 
and 2-4 » broad, which pass over basally without any distinct 
dividing-line into the muscle-fibres. Whether the transition 
is direct, or whether in fact a connective-tissue junction is 
present, cannot be determined from the available preparations 
of Margaritifera vulgaris; but in Mytilus edulis there is a 
distinct connective-tissue layer (fig. 5, ¢.¢.), the fibres of which, 
20* 
[49} 
