A COBAL AND A MOLLUSC. 337 



they are perfectly constant — no specimen having hitherto heen 

 found without such holes, sufficiently proving that a Sipuncuhis 

 lives as commensal in the coral — they here bear the .aspect, or 

 disguise, of a true specific character. 



Many years since, a very interesting case of association or 

 commensalism between a mollusc and a coral was described by 

 Steenstrup. The young of Rhizoohilus antipathum (see fig. 

 88) have all the appearance and characters of a true Buccinum. 

 When they have reached a certain size they attach themselves to 

 the slender branches of a horny coral known as Antipathes, and 

 at once so modify their normal growth that it is quite impossi- 

 ble to name any other mollusc in which any alterations of 

 growth at maturity are to be met with in the least resembling 



Fig. 88. — RMzochilus antipathum^ Steenstmp. To tlie right the yonng shell exactly 

 resembling Bucdnum. ; to the left the old shell firmly attached to the brancTies of AntU 

 ^n/Ae« by its irregularly formed margin. 



it. The shell throws out processes in every direction, by which 

 it clings to the coral, in the mode here shown in a woodcut 

 copied from Steenstrup, till at last the mollusc loses all 

 power of motion and lies, anchored as it were, to the Anti- 

 pathes. Of what use this can be to the animal it is difficult to 

 say ; but we may venture to put forward the hypothesis that 

 it must be, and is, of some service, and also that an originally 

 accidental connection and growth of some true species of Buc- 

 cinum with the slender branches of a Gorgonia must have given 

 rise to this extraordinary habit. 



Certain parasitic Crustaceans ofier another example of such 

 a peculiar action of one animal on another. The species of Pel- 

 togaster often live attached to the hind part of the body of the 

 hermit crab, and they then assume the form necessitated by that 



