248 CORAL AND ATOLLS 



lagoon, algse and other forms become established upon it. 

 Darwin assumed that in these rocks upon the barrier he saw 

 the mark of the limit of coral building to tide-level, for the 

 dead flat top of the rock he regarded as being the result 

 of the exposure of the zooids above low-tide level. Semper 

 also took tide -level to be the cause of this formation, but he 

 added the accumulation of sand, and the occasional presence 

 of fresh water from showers of rain, to the effects of exposure 

 to sun and air. I have given an account of the development 

 of these flat boulders, as I have observed it in the atoll, and, 

 in this account, the exposure to sun and air has been excluded 

 as a causative factor in the great majority of cases. 



A colony of Porites is composed of a number of individuals, 

 every one of which has like powers of division and reproduc- 

 tion ; it therefore follows that when an embryo settles upon a 

 small and isolated nucleus, the colony which it gives rise to will 

 tend to assume the form of a sphere. Such spherical colonies 

 are, as a matter of fact, often to be found lying free in rock 

 pools ; but they can never grow to any size as spheres, and no 

 large colony ever has anything like a truly spherical form. 

 The dome-shaped rock is the type of all larger colonies : it 

 is developed gradually from the spherical type by the death of 

 all the zooids in that part of the colony that is subjected to 

 pressure ; i e., the side of the sphere upon which the colony 

 was lying. A dome-shaped colony may grow to a great size, 

 and some of the largest of all the colonies of the type P. lutea, 

 which live upon the submarine slopes of the outside of the 

 atoll, are very perfect domes. As a rule, wherever sedimen- 

 tation is taking place with rapidity, the dome shape is lost, 

 and the flat-topped rock is produced. Colonies in the lagoon, 

 deep and undisturbed by tide-level, assume this form ; colonies 

 in barrier pools that lie many feet below the lowest fall of 

 spring low tides become flat-topped, and it is evident that it 

 is not the low-water exposure which is responsible for the 

 assumption of this form. 



A dome-shaped colony grows larger and larger, its outline 

 becomes the segment of an ever-increasing circle, and the top 



