1907.] AND SUPPOSED SPECIES IN CORALS. 519 



The modern literature of corals and the modern methods of 

 their study are devoid of interest, except to the museum specialist ; 

 and of those people whose lot is cast in places where excellent 

 oppoi'tunities of study are at hand, many are prevented from 

 making useful observations, by the wonderful maze of man-made 

 difficulties that surrounds the study of a class of really interesting 

 animals. 



So as to introduce some order into the arrangement of the 

 material collected and observed in the Cocos-Keeling group, it will 

 be best to follow the life of the growing coral in its natural 

 surroundings. The early stages need be touched on but lightly, 

 and for all practical purposes the life of the colony may be con- 

 sidered to start with the fixation of the young larva. It is well 

 known that the hollow, and as yet entirely uncalcified larva is 

 expelled from the central cavity of the brooding parent, and in 

 company with many hundreds of its kind is shot oub into the 

 water to take its chances of the tides and currents. It will at 

 once be seen how varied may be the fortunes of the young coral, 

 and how each individual embryo, endowed with its inherent 

 growth-tendency, has a wide range of chance resting-places, that 

 may happen to be suitable or not. 



Every coi-al embiyo launched into the world mvist start its cai'cer 

 by becoming fixed to some nucleus, upon which it may build its 

 future colony ; and the nature of the nucleus, and its situation, 

 though necessarily the outcome of the merest chance, are the 

 determining factors in shaping the after destinies of the colony. 



Coral embryos will l)ecome attached to almost anything in the 

 water ; they become fixed to older colonies, to shells, to dead 

 coral masses, or even to floating pieces of wood ; their resting 

 place may chance to be of almost any shape, and to be situated 

 in almost any environment. Now when the young coi-al starts 

 its division, both the shape and the site of the nucleus are 

 capable of modifying its method of division. 



The asexual reproduction of corals is carried out by the division 

 of the parent zooid, and the various methods by which this 

 division is effected differ from each other considerably ; and it is 

 necessary to touch briefly upon the chief methods by which a 

 zooid reproduces itself. 



It is as well to state at the outset, and thus avoid much further 

 explanation, that the different methods of division, though highly 

 characteristic of certain well-mai-ked types of growth, are not 

 definitely and unalterably fixed ; and I think it is justifiable to 

 dogmatise, and state that any form of coral may exhibit any form 

 of division. Since the resulting form of vegetative growth is 

 purely the outcome of the type of division adopted by the zooid, 

 then it follows that coral colonies may grow in many vegetative 

 forms ; and, again, these vegetative forms are not definitely fixed, 

 for as any form of coral may exhibit any form of division, it 

 follows that any form of coral may also exhibit any form of 

 vegetative srrowth. 



