MARGINAL PROCESSES IN SHELLS. 1059 



I have not certainly traced any damage to them. Tridacna, 

 Spondijlus, and Chama similarly have every chance of reaching 

 old age once the younger stages are passed, and the majority of 

 specimens found are thick and heavy ; on the other hand, Mar- 

 garitifera vulgaris (the Ceylon pearl-oyster), Ostrea sp. ?, and 

 Avicula zebra of a year or two old are common, but above this 

 age are scarcely ever met with in the Red Sea. In the only 

 fishery for Margaritifera vulgaris which I have seen here, shells 

 two or three years old, and small for that age, were being obtained 

 in comparatively insignificant numbers. 



That enei-gy so valuable elsewhere should be wasted upon un- 

 necessary appendages is also negatived by the consideration that 

 in this class, as elsewhere in nature, the struggle for existence is 

 severe, demanding the nicest adaptations to environment. This 

 is shown by the existence in this one bay of three distinct species 

 of Margaritifera, anatomically scarcely distinguishable, apparently 

 living together in the same habitat and obtaining food in 

 exactly the same way. Only after careful study does it begin to 

 be seen what delicate adaptations to special environments keep 

 the three species distinct, each in its own niche in the woi'ld. 



Observations of the relations between living examples of the 

 following species and their common foe, 2Iurex ramosus (text- 

 fig. 229), shows that these oi-namental processes have a simple and 

 essential use to their possessor ; indeed, without them they would 

 fall such easy victims to predatory Prosobranchs that those species 

 not otherwise protected must become extinct. 



The following species living in exactly the same surroundings, 

 in the same aitificial way, were observed, and their mortality 

 from attacks by Murex noted : — 



Ostrea ST^.1 No spinous processes. Decimated. 



Avicula zebra .. . ,, ,, ,, 



2Iargaritifera Processes small and Killed in large 



mauritii. weak. numbers. 



Margaritifera Processes small, except Kept alive by frequent 

 vulgaris. in quite young shells. removal of Murex, 



otherv/ise numbers 

 perished. 

 Margaritifera Processes large and Attacked only excep- 

 viargaritifera. strong, remain well tionally. 



developed to at 

 least 6 3^ears old. 



We see that in these five species the liability to attack by 

 Murex (and probably by other prosobianchs, the behaviour of 

 which I have not observed) is just in proportion to the develop- 

 ment of spines. The cases of attack ujjon M. margaritifera 

 which I have seen corroborated, occurred only in stunted shells 

 in which these processes were ill-developed. 



In the Aviculidse at least the pi-ocesses are made from the two 



