130 MADliKroUAKIA. 



Table III.— ANALYSIS AND DISTRIBUTION OF THE KNOWN VARIATIONS 

 IN OROWTII-FOKMS OF THE POEITKS OF THESE SAME REGIONS. 



Since the preparation of tliu analytical talile on the jirowth-f'orms of the Indo-Pacific Porites 

 in Vol. v., and even since tlie printing of the .systematic ilescrijition in thi.s Volume, some 

 further insight has been gained as to tlie general principles of growth in the Stony Corals — see 

 for detftils, Introduction, Sections III. and IV. We nevertheless still .start from the same 

 ideal initial colony, a small plano-convex disk, for from it all the later forms may be deduced, 

 not as wo assumed in Volumes IV. and V. by dii-ect continued growth, but by a succession of 

 repetitions. The initial colony aj)poars to grow continuously to a certain size, and during tliis 

 growth it assumes some definable shape of its own. Its further growth as an individual is 

 then ajiparently arrested, and by means of some asexual reproductive process not yet uudei-- 

 stood, the colony repeats itself and a new colony appears upon the old. 



The shapes assumed by the initial colony may be roughly divided into two : those which 

 grow mainly in width and those wliich grow in height. All those which grow in width 

 necessarily build massive stocks by the repetitions of their colonies, each new colony covering 

 over and killing the one from which it sprang. Tlie new colonies of those which grow in 

 height on the other hand need not necessarily each time kill the parent, at least as far as we 

 can see. Although, if they did, we should therein find a reason for the phenomenon that many 

 tall forms have the gi'owing colony confined merely to their tips. Hitherto, it has usually 

 been customary to refer this to the crowding of the stems. But while that is a good cause in 

 many cases, it will not always apply. 



In all former tables we have taken the shape of the whole stf)ck alone into consideration : 

 it is clear that now in all future classifications of growth-form, we shall have to find the .shape 

 of the colonial units of which each stock is composed. It is these, one would think, which 

 should supply us with our basis of classification. So Little, however, is yet known of them 

 and of the degrees to wliich they may be specialised and disguised, that we propose to continue 

 in the old paths, merely calling attention to any clear cases of metameric gi-owth which we 

 see. In the future, however, there can be no doubt that the metameric factor in stockbiulding 

 will have a definite place in systematic descriptions. I say a definite place, because it has 

 already uncon.sciously forced itself upon the notice of systematists, e.g., in such cases as tiiose 

 in which a massive form is obviously built up of successive layers, and is variously described 

 as massive, and as encrusting, according as the mind of the student is most attracted by the 

 mass or by the successive layers of which it is built. 



It must indeed be admitted that it is even now difficult to decide how such forms should 

 lie classified. At present we can do no better than take the form of the stock, rather than of 

 the individual segments. 



A. riano-Gonvcx. 



This ideal primitive colony, which plays such a large part in building up the stocks of 

 Gmiiojmra, has, as already noted in Vol. V., p. 258, not persisted so fre(|uently in Porites. One 

 explanation of this may be that simple undifferentiated Porites colonies wo\ild be much 

 smaller than those (jf Goniopora, and therefore have escaped the vigilance of collectors. Stocks 



