546 DR. F. AV. JONES ON GROWTH-FORMS [June 18, 



course of 100 days, whilst larger areas were commonly attacked 

 by sponges or algse before repair was completed. Very much the 

 same process is seen to occur when constricting metal bands are 

 placed round branches. Tlie zooids, where the bands exert theii- 

 pressure, are destroyed, and the general growth of the branch and 

 the sympathetic activity of the zooids at the margin of the injury 

 soon tend to cause the band to sink into the substance of the 

 coral, and to become completely embedded beneath the surface. 



It was pi'eviously stated that the type of repair naturally tends 

 to follow the type of growth of the injured colony ; and this is 

 generally ti-ue. If a coral has a particular mode of growth, and 

 if it be in such an environment that its mode of growth is the 

 one most suitable, then an injvuy is repaired or a part regenerated 

 on the same type of growth as that of the colony. But the par- 

 ticular form in which a colony may be growing may not be the 

 form best suited for successfully flourishing in that particular 

 environment ; and then, following an injury, a remarkable state of 

 things is brought about, which may be stated as follows : — 



II. A coral may repair an injitry hy a new groivth of a different 

 type to that of the colony^ and in such cases the repair gro'wth 

 is of a form better suited to the environment than is the form 

 of the p>arent colony. 



It must be borne in mind that the physical conditions of the 

 atoll are not absolutely constant ; seasonal changes in sets and 

 strength of currents are perpetually causing alterations at all 

 points of the island ling, and a habitat may alter greatly in its 

 physical conditions in the course of a few months. When a coral 

 embryo settles down upon a basis, and starts the fovinding of a 

 colony, the type of growth that wdll result will be what is best 

 fitted to the environment as the embryo finds it. But suppose 

 that, after the founding and growth of a colony, the physical 

 conditions of the environment change, rough water is admitted, 

 silt settles down, or the water becomes shallow and calm : then 

 the growth of the colony may not be the ideal one for flourishing 

 under these new conditions. 



In many places where the conditions are prone to vary, the 

 habitat of the corals may alter in its physical nature in more or 

 less regular cycles. Where, as at the eastern extremity of Pulu 

 Tikus, a spit runs out, building sheltered pools and protecting 

 the shore for a considei-able distance, and then late^r on, in the 

 periodical cycle of currents and eddies, is carried away; many 

 corals must be subjected to a great variation of envii'onment. 

 In such circumstances no doubt many colonies die, for, as we 

 have seen, a sudden change of habitat cannot be resisted wdien a 

 form is well adapted for one definite kind of environment, and in 

 any case the colonies are liable to injury and partial death in 

 their changed surroundings. 



When such injury falls on a coral not ideally situated, the 



