1907.] AND SUPPOSED SPECIES IN CORALS. 525 



an exact impression of every irregularity of its contour; and 

 then, reaching the limits of the basis, they may grow from it as 

 flat plates, spreading out from its margins. Such corals as grow 

 in this fashion show many curious changes at their free margins, 

 for although the superior surface of the encrusting layer can 

 alone produce zooids, still at the free edge zooids will appear on 

 both superior and inferior surfaces. 



In connection with these partial plates a very curious fact is 

 always demonstrated, for the whole surface structure of the coral 

 body, and of the corallites, differs above and below ; and this 

 important fact will need further reference. A coral may start 

 from its first beginning by growing as free exfoliating plates, and 

 then it may bear zooids on both surfaces of the plates, or on the 

 upper surface alone. In every case where zooids are borne both 

 above and below, there is the same marked difference of structure 

 between the two surfaces. 



Text-fio-. 152. 



Vegetative reproduction in Corals. 



Diagrammatic section to illustrate the mode of growth of Montipores : 

 the uppermost zooids are the most newly formed. 



When the linear growing point grows unifoi'mly upwards, the 

 I'esulting growth consists of a series of vertical plates ; and when 

 the growth takes this form, the structure of both sides of the 

 plate is identical, and zooids grow from both surfaces. 



A plate-like growth is formed by a uniform and continuous 

 growing edge ; but. the linear series of young growing cells may 

 not maintain their continuity : the growing edge may reach a 

 certain size and then divide, and the resulting growth consequently 

 takes the form of a plate, cleft at its edges, or of a branching 

 form, all of whose branches are given off in one plane. This is a 

 highly chai'acteristic form of growth of one type of Montipore 



