1902.] ORIGIN OF PEARLS. 155 



small simple Sporocysts, or, at most, groups of three or four 

 secondary ones. Examples which averaged 17-21 mm. in length 

 contained masses of seven to ten cysts, those measuring 27 mm. 

 had still larger groups, while from that size upwards there was a 

 steady progression in the dimensions and number of constituent 

 units in the groups of Sporocysts. Some of those found in Tapes 

 40-50 mm. long measured 7 mm. in diameter, and contained as 

 many as thirty secondary cysts, and a hundred Oercarite or even 

 more. 



As the Sporocyst grows, it sometimes compresses the tissues 

 that intervene between it and the shell, which apparently inter- 

 feres with the secretion of fresh shell-layers. This leads to the 

 development of white chalky spots on the inner surface of the 

 valve. These patches, in old individuals, further prove that the 

 large groups of Sporocysts are the descendants of the original 

 small ones which are present when the molluscs are young. For 

 we may see two or more white scars on the lining of the shell, 

 mai'king the position of the cysts when the shell was younger. 

 In fact, the Sporocysts may leave, imprinted on the shell, the 

 history of their movements as the mantle-margin shifts outwards, 

 just as the adductor muscles mark on the nacre the record of their 

 migrations. The first or innermost of these scars is a small one, 

 such as might result from a triple or quadruple cyst, the next 

 is larger, while the group of Sporocysts in the mantle-margin is 

 larger again. This plainly shows that as the mantle-margin 

 followed the peripheral growth of the shell, the group of Sporo- 

 cysts increased in size. 



These compound Sporocysts are, of course, thicker than the 

 normal thickness of the mantle, and stand out as opaque white 

 granular eminences, obvious as soon as the shell is opened. 



On the beds of Pearl-bearing Mussels in the Barrow Channel, 

 opposite the Piel Fish- Hatchery, where every specimen of Mytilus 

 is abundantly infected with the Leucithodendrium, and almost 

 every specimen contains pearls. Tapes is not found. The Cockle, 

 Gardium edule L., is common there, and acts as a host for the 

 Sporocysts. Somewhat less than half the specimens of Cardium 

 that I examined at Piel were infected. In Gardium the Sporo- 

 cysts occurred in the mantle-margin, close to the anterior border of 

 the anterior adductor muscle. Large groups, such as occur in 

 Tapes, were not observed, but only single, triple, or quadruple 

 cysts. 



I have not yet been able to trace the infection of Tapes or 

 Gardium. It is therefore impossible to say whether infection 

 takes place by means of a free swimming Mi7'aGidiu,m larva or not. 

 The constant occurrence of the Sporocysts in exactly the same 

 positions suggests that the eggs are carried into the digestive 

 system of Tapes with the food-bearing current, hatch out in the 

 alimentary canal, enter the circulatory system, and reach their 

 destination via the posterior pallial artery, along the course of 

 which they are distributed in Tapes. Moreover, the position in 



11* 



