PORITES. 21 



up of colonial units. Indeed in the active formation of any coral stock we have three units of 

 structure ; cells build up the calicles, the calicles build up the colonies, and the successive 

 colonies build up the stock. We are here concerned with the two latter units alone. 



The Stony Corals then exemplify the growth-principle called by zoologists " metamerism." 

 It differs from ordinary gemmation, which means budding generally, in that it is serial ; and 

 it diff"ers from sexual reproduction in that the latter begins the new organism de novo from 

 sexual cells, whereas gemmation of one multicellular organism from another is, I take it, a 

 very specialised form of fission, in which the area affected has been secondarily limited and 

 circumscribed, so as not to cripple the functions of the parent so completely as true fission — 

 that is, fission into two equal parts — would do. 



In metameric stock building the " budding " of the units is serial, and the following 

 conditions might occur : — 



1. The parent — which, in the corals, may be a single adult or a colony — having budded 

 may die down, and the total result is a succession of organisms piled up on one another, the 

 last of wliich is the only one living. 



2. A second, higher stage would be reached when the parent forms die away more 

 slowly and progressively, so that large stocks are produced in wliich an indefinite number 

 of the more recent units are alive at one and the same time. 



3. A third stage would be reached when these living units become co-ordinated together 

 for mutual aid and common action, defensive and offensive ; while lastly — 



4. A fourth stage would be reached when the generations cease to appear as separate 

 generations, but merely as segments, and even vestiges of segments, and what was really a 

 succession of asexual generations becomes as one organism. 



With reference to these stages, it is clear that the early death or continued life of the 

 parents must largely depend upon the shape of the organism ; if, being sessile, it is flat like a 

 plate, the bud is bound to smother the parent ; if it is of any other shape, the parent has some 

 chance of further existence ; while if the organism is free-swimming the whole colony might 

 live and move as one organism. 



These principles can be shown to have played a great part in the building up of the 

 larger living organisms, and in many it is still traceable in the result. In the vegetable 

 kingdom we point to the trees and the grasses as forms obviously built up by the repetition 

 of parts. In the animal kingdom we may point to the Polyzoa, and to the worms with 

 their many segmented derivatives, Insecta, Annelida, Crustacea, and, as some think, the 

 whole of the Vertebrata. In the first three of these latter we find all stages in the gradual 

 obliteration of visible segmentation with increasing concentration of function. The animal 

 gioup which, however, here most interests us is that of the Coelent«rates, to which the 

 Stony Corals belong. 



In this primitive group, growth by gemmation is one of the most striking methods of 

 reproduction ; there are many variations of it. The buds may adhere and fonn branching 

 stocks ; some may break away as sexual individuals. Others build up free-swimming stocks, 



