PORITES. 23 



normal size of the adult is reached the growtli is diftei'ent, and can no longer be described as a 

 growth of the individual with withdrawal (jf the enlarging polyp from too small a skeleton, 

 but the dejjosition of a new skeleton upon the framework of the old, the new skeleton 

 belonging to a new individual polyp which comes into existence in some way by gemmation 

 from the old, but the only visible sign is still seen in another series of tabulce, differing from 

 the first in that there is now no increase in the diameter of the stock. 



We find, then, in these so-called solitary corals, that they are also in reality colonies — pro- 

 duced by gemmation — in which, however, only the last member of tlie series remained alive. 

 This explanation so far appears fairly satisfactory : at the same time this is not the first time 

 the phenomena they present have been discussed, but it seems hitherto to have been mainly 

 attributed to some peculiar method of rejuvenescence (" verjiingung-process"*). If the linking 

 on of this process with gemmation, as here suggested, will stand the test of further inquiiy, 

 that term will hardly be necessary. 



From these solitaries we pass to what are ordinarily known as colonial corals, and we 

 find distinct traces of the same metameric growth. These colonial forms must therefore be 

 regarded as the result of two distinct kinds and periods of gemmation or colony formation. 

 Let us suppose that one of these large flat saucer-shaped calicles of Paleozoic times divided up 

 into a colony of small calicles ; the stock would then consist of tiers of small disk-like colonies, 

 and we should get an almost exact reproduction of what we actually find in Kfe, e.g. in the 

 genus Goniopora. Indeed without having perceived the meaning of the phenomenon so clearly 

 as I do now, I have already shown f the 'apparent homology between the metameric re- 

 duplication of single calicles such as is seen in Montlivaltia with their successive bands of 

 epithecal films (cf. the tabulfe), and the succession of growth periods in Goniopora with their 

 closely similar bands of epitlieca, which are again obviously tlie edges of tabular (see Vol. IV. 

 of this Catalogue, p. 24, Diagrams A, B, C, D). In this last named genus we have the growtli 

 method here discussed exemplified in a very striking manner ; for, on attaining a certain size, 

 the initial Gonioporan colony is always grown over by another, with a distinct edge, and conse- 

 quently with its own epithecal rim. 



The whole phenomenon has been a standing difficulty to me, and until now I have seen 

 no solution, although I have puzzled over it many times. It was especially difficult because 

 each colony was apparently grown over by another, and that by another, into a series of 

 colonies one on the other. This was obvious from the many distinct edges, yet a section 

 showed the whole series continuous throughout, the trabeculce running smoothly through the 

 whole. 



How, I thought, is it possible to have a new and distinct colony with its skeletal elements 

 continuous (except, perhaps, at the edges) with those of the skeleton on which it grew ? We 

 can now say that it is only another case of the continuity of the skeleton of the bud with that 

 of the parent, such as was described above for the fossil solitary corals. The matter is here, 



* Milaschevitch, Palaeontographica, xxi. (1876) p. 223 ; Tomes, Geol. Mag., 1888, pp. 208-9 

 I Ann. and Mag. Nat. Hist., 7" xiii. (1904), p. 9. 



