THE CEYLON PEARL-OYSTER. 329 



PI, XL. fig. 2G .shows ;i penrl having for its nucleus a double 

 hypostraouiu-peai'l, the two constituents being apparently 

 separated b}'^ a mass of amorphous sul)stance. Other examples 

 were examined which showed iutei-gradations between double 

 and single hypostracum-pearls. 



in iig. 27, from the same preparation as the above, there is a 

 central cavity, containing a few granules, and surrounded by 

 amorphous and columnar substances followed by nacre and finally 

 by hypostracum. In another case in the same preparation the 

 same state of affairs was shown, but there were further layers of 

 nacre outside the hypostracum, resulting in a fair-sized pearl. 

 These cases are of interest as indicating that the attachment of 

 muscle-fibres to a pearl may arise secondarily, in the coui-se of 

 its growth. In some cases the columnar substance is stratified, 

 as in the pearl at the bottom left-hand corner of fig. 19, 

 PI. XXXVI [I. and in fig. 49, PI. XLIY. In PI. XL. fig. 28, which 

 is the left-hand pseudo-nucleus of the double pearl shown in 

 fig. 49, there is at one pole a distinct transition from the columnar 

 matter of the pseudo-nucleus to normal nacre {nac.'), just as 

 occurs in the substance of the pearls figiu-ed at Pi. XLI. fig. 34 

 and PL XLIIL fig. 43. This pseudo-nucleus is followed by 

 normal naci-e (^nac), after which there is a layer of brittle horny 

 matter (am.), which I regard as analogous to amoi-phous repair- 

 substance ; this is followed by another columnar layer (col."), 

 imperfect on one side, and this again by nacre. 



A preparation from the second unlabelled specimen in the 

 British Museum showed, side by side, all kinds of intergradations 

 between the various kinds of nuclei and pseudo-nuclei of muscle- 

 pearls described above. 



I regret that I am as yet unable to ti'ace the origin of the 

 muscle-pearls beyond the cyst-like sac, lined with hypostracum 

 or another form of shell-substance. One or two preparations 

 show cavities lined with a few cells, which may represent the 

 first stages of pearl-sacs. Moreover, here and there, the relations 

 of the epidermis, at the place where the muscle-attachment 

 epitheliiuu gives place to ordinary mantle-epithelium, suggest 

 that these cysts arise by direct invagination at this region. What 

 may be the mechanism which gives rise to this invagination I 

 cannot say, there are certain indications that suggest that it may 

 be associated with the presence of parasitic Protozoa, an idea 

 suggested by me in 1902 (25, p. 162) ; but until I have been able 

 to examine better-preserved material, it would be unwise to 

 hazard yet another theory of pearl-production. At any rate, it 

 is clear that, if pearls in 31. vulgaris ai-e of parasitic origin (and 

 T adhere to the view that the distribution of pearl-producing 

 examples of this and most other molluscs can at present best be 

 explained on the parasitic hypothesis), the immediate cause of the 

 pearl is not the mechanical irritation caused by the body of the 

 parasite, but rather the toxic prope7-ties of its secretions, which 

 lead to the pathological changes (formation of the tumours that 



