530 DR. F. W. JONES ON GROWTH-FORMS [June 18, 



reproductive value of all zooids tends to the production of a 

 sphere, it is of course the normal habit of growth ; and so such 

 corals have their natui'al home in the rougher water, and undergo 

 their atypical modifications when developed in other sites. 



All branching forms, wdien growing in rough water, are sub- 

 jected to a change of habit of vegetative growth ; and this change 

 is a very interesting one. It will be pointed out when the process 

 of repair is dealt with, that injury aflects various forms of corals 

 differently ; and all these rough-water forms are in reality the 

 result of perpetual injury. 



It would be an obvious disadvantage to a coral to adopt a highly 

 branching form, when living on the surf -beaten portion of the 

 barrier ; for the colony would soon be wrecked and broken up by 

 the waves. There is no Montipore in the atoll that lives by 

 election in rough water, bvit occasionally colonies become exposed 

 to strong currents ; and then the perpetual injury to the upper 

 portions causes the growing clusters to be bi'oken up and confused ; 

 and the resulting growth is an irregular mass of short stunted 

 branchlets. 



Montipores in rough water may also take on a creeping habit 

 of growth, and form encrusting layers on the surface of dead 

 massive colonies ; but they are not corals that are at all common 

 in any but the calmest water. Their usual home is the lagoon, 

 and the greatest normal depai'ture from their favourite habitat is 

 in the current-swept shallow inlets to the lagoon ; and here an 

 endless series of modifications may be collected, ranging from 

 branching forms to mere amorphous masses and encrusting 

 layers. (PI. XXVII. fig. 1 c.) 



The rough-water forms of the Madrepores are highly chaiac- 

 teristic, and all depend on the processes that always occur in this 

 gTOup when the " dominant apical zooid " is injured. Instead of a 

 growth that consists of few dominant zooids situated at the 

 extremities of long branches, a rounded mass is formed ; and it is 

 composed of little groups of symmetrical zooids surrounded by 

 others that are asymmetrical — each little group representing a 

 blanch in an extremely abbreviated form. 



From these exti'eme rounded masses, wdth branches that are 

 mere bosses on the general sui'face, every transition foi'm may be 

 collected, up to the highly developed lagoon types, with branches 

 many feet in length. 



Exactly the same results are produced in the Pocillopores, and 

 here the rough-water type is a flattened growth, w^ith irregular 

 divisions into separate lobes ; each lobe representing a separate 

 branching system (Plate XXYII. fig. 3 c). 



Any coral that chances to establish its colony in rough water 

 exhibits therefore one well-marked characteristic : it always tends 

 to form a rounded, or flattened mass ; and this for obvious 

 mechanical reasons. It is a direct outcome of the conditions of 

 the environment, and any zooid must conform to it or peiish. 

 Repeated injury to the growing ceils is the detei'mining cause of 



