324 DR. H. LYSTER JAMESON ON 
80 p in diameter, and has a wall about 10 » thick, lies close to a 
nacreous muscle-pearl, about 1 mm. in diameter, ‘ne sae of which 
is shown at ep.p.s. The cyst is embedded in a strand of muscle 
traversing the mantle-parenchyma obliquely, and ending in muscle- 
attachment epidermis which was attached to the shell. (Such 
connections between the general musculature of the mantle and 
the shell occur here and Pere quite apart from the more regular 
muscle-scars. For examples of this in J/ytilus see List, 27 b, 
Pl. 8. fig. 1.) The cyst contains at one point a little ereralet 
mass. ‘The muscle-fibres here appear to be in direct contact with 
the hypostracum. The easiest explanation of this condition would 
seem to be the hypothesis that the original epithelium has dis- 
appeared. It is not difficult to suppose that a highly specialised 
“tendinous ” epithelium, like the attachment-epidermis, whose 
fate seems to be to become a part of the shell, is incapable of 
regenerating itself, and, therefore, destined to die and disappear 
on ceasing to be functional. If we take this view, the typical 
hypostracum- -pearl is not so much a stage in the dev elopment of a 
nacreous pearl as a phase parallel with it; the latter arising when 
the original sac contains some of the ordinary n nacre- secreting 
epidermis, or cells capable of giving rise thereto, the former when 
it is composed of attachment- epithelium alone. The hypostracum- 
pearl would thus have a limited growth, the nacreous pearl an 
unlimited growth. However, in considering these cases where 
there does not appear to be any attachment-epithelium, it must 
be remembered that this particular epithelium is often very difficult 
to see, so that some workers have even failed to detect its existence 
on the regular muscle-insertions. Much light can no doubt be 
thrown on these questions by a really thorough study of the 
behaviour of the cells at the places where the muscle-attachment 
epithelium goes over into the ordinary epidermis of the mantle, 
and of the histological phenomena associated with the wandering 
of the muscle-attachment. The material of the pearl-oyster that 
T have examined so far is not sufficiently well preserved to allow 
of such study. So far as I know, this important matter has never 
been properly investigated in any molluse. 
T will now pass from the hypostracum muscle-pearls to the 
nacreous muscle-pearls. ‘Typical instances of these are shown in 
Pl. XXXV. figs. 8&9 and Pl. XXXVI. fig. 10. These three 
examples are all explicable as derivatives of the hypostracum- 
pearl. Figs. 8 & 10 obviously lie in the borderland between 
one of the ‘regular muscles and the parenchyma (fig. 8 is at the 
insertion of one of the pedal levators). Fig. 9, from one of 
Prof. Herdman’s slides, is in a place in the free mantle where a 
few small muscle-strands (musc.) are attached to the shell. The 
sac of each of these pearls is lined in part by ordinary nacre- 
secreting epithelium, underlying which is the typical granular 
parenchyma, in part by muscle-attachment epithelium, continuous 
with the musculature. As the former is much more active than 
the latter, these pearls are all eccentric in shape, having a hilum 
[66] 
