Since the three superimposed layers of the shell are secreted by separate parts of the mantle, _viz._, the nacre by the general surface, the prismatic layer by the inner edge, and the epidermis by the outer edge, it follows that if a pearl in course of formation is moved from one of these distinctive portions of the palial organ to another, the nature of its laminæ changes. Thus, if a pearl formed on the broad surface of the mantle is moved in some way to the inner edge of that organ, it may be covered with a prismatic layer; if then moved to the outer edge it may receive a lamina of epidermis, and then by changing again to the broad surface of the mantle it receives further coats of nacre. [Illustration: Pearls from common clam (_Venus mercenaria_) of eastern coast of America ] [Illustration: Pearl “nuggets” from the Mississippi Valley ] [Illustration: Wing pearls from the Mississippi Valley ] [Illustration: Dog-tooth pearls from the Mississippi Valley ] The structure of pearls from univalve mollusks, such as the conch, the abalone, etc., as well as those from some bivalves, as the Pinna, for instance, differs from that of the true pearls formed in species of Margaritifera. Instead of the alternate layers of conchiolin and of carbonate of lime, many of these have an alveolar structure. When greatly magnified, the surface of a Pinna pearl appears to be formed of very small polygones, which, as decalcification shows, are the bases of small pyramids radiating from the nucleus. The walls of these pyramids are formed of conchiolin, and they are filled with carbonate of lime of a prismatic crystalline structure. This is simply a modification of the parallel laminæ in the Margaritifera pearls, for, as Dubois points out, in some sections we can see portions where the alveolar formation has proceeded for a time coincidentally with the lamellar form.