THE CKVLOX TEARL-OYSTER. 301 



layer of tlie periostfacum is closely adherent to the specialised 

 e[)itheliuiii, so that secondary thickening, if it takes place (and 

 niy own observations on Mytilas and Modiola lead me to think 

 that it does so to some extent, a view v/hich is held also hy 

 List, 27 h, [). 55), differs from that of all other parts of the shell 

 (inclnding the inner layers of the periostracnm) in that it 

 is seci-eted from ontside or centrifngally with respect to the 

 body of the animal and tlie shell, instead of from inside or 

 centripetally. It would, indeed, seem as thongh there were 

 morphological gronnds for restricting the name periostracnm to 

 this particular layer of the outer cuticle-like snl)staiice, or, failing 

 that, for introducing a term which would separate it moi'e sharply 

 from the more bulky inner layers. The structui'al distinctions of 

 this layer ai-e well shown by Riimer (32) fig. 25 (Jfargaritana), by 

 Moynier (28) tig. 50 (Mj/tilus), by Tullberg (47) Taf. iv. fig. 3 c, 

 lig. id (Mi/tilus), and by List [27 b) in the Mytilidpe generally. 



The greater pai-t of the periostracnm, however, is laid down 

 centripetally, layer upon layer, by the epithelium on the outer side 

 of the marginal gi'oove (v. e. on the inner siile of the outer of 

 the two lobes of the mantle-margin, Tullberg (47) p. 27). It is 

 stratified, and in Ih/tilus contains, near its outer limit, a charac- 

 teristic layer of large alveoli. The stratifications of this substance 

 have been shown by Romer to correspond to lavers of minute 

 aiveoli (32, fig. 25). ' 



Internally the inner layers of the periostracnm pass over into 

 the conchyolin framework of the prismatic layei'. 



In some forms, e. g. Auodonia, the distinction between the two 

 constituents of the periostracnm are emphasized by the outer 

 layer being much greater in area than the inner ones, and being 

 thrown into folds upon which the inner layers lie unconformably. 



The Mother-of-Pearl Oysters (together with such forms as 

 Ostrea and Pecten) difter from the types to which the above 

 description refex'S in the fineness of the periostracnm and in having 

 much more freely I'eti'actile mantle-mai-gins. In sections of the 

 decalcified shells of the Mother-of-Pearl Oysters it is difficult — 

 indeed, I might say impossible — to diftei-entiate the periostracnm 

 from the outer layer of the organic basis of the prismatic sub- 

 stance (text-fig. 35, p. 302 ; see also PI. XXXIV. fig. 6). Here 

 mai-ginal gi-owth proceeds by a series of steps and retreats, the free 

 mantle-margin being retractile to the edge of the nacre, and being 

 so retracted when the shell closes. Thus, after a new process 

 of the lip has been formed, the mantle-mai-gin is withdrawn, 

 and forms a fresh attachment on the inner sui'face of the last- 

 foi'med lip, from which a fresh lip is pi'oduced. So the periostracnm 

 of lip no. 2 is attached to and apparently a direct continuation of 

 the inner surface of the conchyolin of the prismatic layer of lip 

 no. 1 , and constitutes the outer layer of the conchyolin of the 

 prismatic layer of lip no. 2. But Herdnian has shown that, 

 difficult as it is to distinguish a. separate periostracnm in sections 

 of the shell, such a layei', of extreme delicacv, does exist at the 



