MADREPORARIAN CORALS 29 



polyp and calyx from the margin of the calyx of the youngest 

 terminal calyx of the branch, followed in turn by the forma- 

 tion of a new polyp and calyx on the opposite margin of the 

 former when it has reached a later stage of growth. This 

 alternate right and left budding gives the younger branches 

 of a large colony a curious zigzag appearance, but in the 

 older branches the angles become smoothed out by the 

 continuous growth of the coenosteum until they appear 

 to be perfectly straight. 



The study of this method of growth in Lophoheha is 

 necessary in order to understand that the calyx of a coral 

 corresponds with only a part of the corallum of a soHtary 

 coral like Caryophylha. The lower part of the theca of 

 the solitary coral is represented by that part of the branch 

 which extends from the calyx to the level of the next calyx 

 on the opposite branch. In other words, it might be said 

 that this colony is formed by the coral polyps growing on 

 one another's shoulders. 



In other colonial corals it is not easy or not possible to 

 demonstrate that the colony has been formed in this way, 

 and consequently the coenosteum or matrix which bears 

 the calices appears to be entirely communal in origin and 

 function. 



In another colonial imperforate coral, Ociilina, for ex- 

 ample, which is usually placed in the same family as Amphi- 

 helia, the calices do not project above the general surface 

 of the coenosteum, and are not arranged alternately right 

 and left, but seem to be arranged spirally or scattered about 

 irregularly on all sides of the branches, and when the branch 

 is examined in transverse section there is no axial series of 

 chambers such as we find in the Lophohelia branch. 



In the family Astraeidae or Star corals, which are usually 

 placed next in order to the Ocuhnidae, the cahces are more 

 crowded together, so that the amount of coenosteum between 

 them is reduced to small dimensions (Galaxea), or the 

 calices come into such close juxtaposition that there is little 

 or no coenosteum at all. 



Before passing on to the complex forms of corallum 

 produced by complete and incomplete fission or by perfora- 



