120 CORAL AND ATOLLS 



It is common, too, to find that the repair-zooids have raised 

 corallites, when those of the parent colony are flat, and in these 

 cases it is probably silt that has caused the original damage. 



The study of the repair in corals is therefore one not 

 devoid of zoological interest, for it shows clearly that a type 

 must never be considered as a species, in the way in which we 

 regard species among the higher animals, until it has been 

 seen in all its variations, and until all the possible modifications 

 that repair produces have been studied. 



That a type like the encrusting Montipo7U should be in 

 reality the same species as the branching form would be 

 considered as highly improbable, but when it is seen that the 

 one type repairs its damage by the development of a new 

 growth of the other type, there is no alternative but to regard 

 them as identical species. 



That the numerous types of Millepora and of Pocillopora 

 should be but variants of a single species would seem at first 

 sight to be very unlikely, for there is little enough likeness 

 between the extreme forms, and yet their processes of repair 

 show them capable of building to any of their diverse types, 

 regardless of the nature of the parent growth. 



If the processes which have been described, and the con- 

 clusions which have been drawn from them, be accepted, they 

 can serve only to make clearer the great fact that morphology 

 — the animal's type — is the outcome of necessity ; and here 

 the demands of necessity bring about change, not in the life- 

 history of a species only, but in the life of the individual. 



