FEEDING OF CREPIDULA. 475 



In connection with the feeding habits of whole chains, an interesting 

 adaptation has many times been observed : when a number of chains 

 have estal:)h'shed themselves on a surface of limited extent, as on a valve 

 of an oyster-shell, the chains are to be found witli their anterior ends 

 towards the edge of the oyster-shell and their posterior ends all con- 

 verging on the middle. Thus the animals are again found to be 

 making the best use of the space at their disposal for ensuring 

 an equally good respiratory and food-current to each member of 

 the group. 



From a consideration of the mode of feeding of Crepidula its habits 

 are easily explained, and it is now easy to understand why all indi- 

 viduals but small ones run a great risk of death if they become dis- 

 enchained, or detached from their surface of attachment. A detached 

 individual with its mantle cavity exposed is unable to produce an 

 effective respiratory and food-current, which along with some un- 

 known cause which gives rise to the general discomfort usually 

 exhibited eventually results in death. Moreover, if a new surface of 

 attachment is offered, it is necessary for the animal to be able to fit 

 its shell fairly well to it in order to have a chance of living. If, 

 however, an animal be offered a surface to which it is able to accom- 

 modate its shell, I find that it is usually able to accept it, especially 

 if the surface offered is smooth. 



XVL THE MANDIBLES OF CKEPIDULA. 



The location of the mandibles of Crepidula appears to have given so 

 much difficulty that even Troschel confesses (18), " Die beiden Kiefer 

 habe ich nur einmai bei Crepidula fornicata gef unden. Sie Konnen also 

 leicht iibersehen werden. Der eine Fund beweist dass sie vorhanden 

 sind, und dass die Meinung diese Familie sei kieferlos welcher ich lange 

 Zeit gehuldigt hatte, irrthumlich war." 



During the foregoing research, however, a clue was given from the 

 mode of feeding as to the probable position of the mandibles, so that it 

 was possible to make a preparation to demonstrate them at once. As 

 Crepidula in eating takes food into the mouth between the roof of the 

 mouth and the dorsal surface of the ladula, one would expect to find the 

 mandibles in the dorsal wall of the buccal cavity. If this region be 

 exposed after dissecting out the radula and its muscular apparatus, the 

 mandibles are to be found lying transversely on a prominence near the 

 mid-dorsal line and just behind the mid-dorsal anterior edge of the 

 buccal cavity as in Fig. 20. They are easily found by following the 

 above directions after soaking the head region in glycerine. 



