20 MADlM'll'OkAKIA. 



network of which thu calicle skeleton may, for practical purposes, be regarded as a 

 modification. 



One other factor deserving of note in the West Indian forms is the obscurity at first sight, 

 of the septal granules (sg of the diagrams). They seem to be frequently confused with the 

 zigzag walls, or else to be but slightly developed. In specimens of the astrteoid * group they 

 can be frequently only just detecied as faint thickenings of the short septa (see, however, the 

 treatment of these granules in Table IV.). 



There are, however, a few cases in wliich the characteristic zigzag is replaced by thicker 

 reticular walls. Both kinds can often be seen on the encrusting and massive forms. On 

 surging-up mounds the walls are thick ; in the valleys they are thin ; while again, others 

 are thick all over, e.g. P. Belize 3 (PI. I. fig. 7) ; P. West hidies x. 15 (PI. V. fig. 3). 



One observation deserves mention, viz. that when foreign objects come too near the 

 surface of a colony and the natural functions of the polyps are interfered with, the walls tend 

 to proliferate. P. Bahamas 1 shows two patches : one a perfectly flattened area due to the 

 proximity of some foreign surface (in this the wall has proliferated in flakes, see PI. IV. fig. 3) ; 

 and a second, where the polyps of a Mussa came into inconvenient proximity with those of 

 the Pontes, with the result that the normal skeleton of the latter is changed from that shown 

 in PI. IV. fig. 2 to PI. IV. fig. 4, in which the proliferation is finely filamentous. It is certainly 

 histructive to find on one and the same specimen two such different results due to the incon- 

 venient proximity of foreign bodies and organisms (see further Table IV.). 



IV. FURTHER EVIDENCE OF THE 

 METAMERIC GROWTH MENTIONED IN THE FOREGOING SECTION. 



In the last section we found traces in Porites of a principle of growth not hitherto recog- 

 nised in the Stony Corals. It would not be true to say that no instance of it has been known, 

 for, as we shall see, the most prominent instances of it have already attracted attention and 

 provoked discussion. Nevertheless, the principle as a principle has not hitherto lieen 

 recognised, and consequently, its significance for the understaudins of the general morphology 

 of the Corals has been entii-ely missed. 



The ])rinciple is tliis, that each stock begins with its larval parents, and by gemmation 

 pnjiluces its initial colony and assumes some characteristic shape. This initial colony is 

 incapable of indefinite gi^owth, but at a certain period, differing doubtless with each form, 

 undergoes some process of repetition, further details of which will be mentioned below. 

 Just as the initial colony was built up by the addition of calicle units, the adult stock is built 



• Used in the sense given on p. 142. The term, always in this Volume, refer.s to the character 

 of the calicle, and i\ot to the growth-form ; although, as far as 1 know, such calicles are only found 

 on massive forms. 



