156 MADREPORARIA. 



encrusting, and the calicles are those typical of that growth-form ; see Vol. IV. Table TVa. 

 p. 176. But on examination of the section, parts of it are seen to be continuous with the remains 

 of a thick mass of the same growth. This mass is, however, burrowed through and through, 

 apparently by a boring sponge, and is quite rotten throughout. This destroying organism 

 follows up the growing coral and burrows just below the living polyps, making gangways and 

 slits, the roofs of which appear to be tlie natural under-surface of a creeping colony. Indeed, 

 patches of the living layer were apparently killed, and the destroying organism reached the 

 surface, but the coral seems to have been able to creep over again. In this way, its method of 

 growth, owing to its perpetual struggle with the sponge, is one of continual encrusting over the 

 parts burrowed through, and its calicles are characteristic of the encrusting method of growth. 



Since the two chief characters of this coral are due to the action of the sponge, its specific 

 affinities must necessarily be left until the Adolphus Island reefs are again explored, for until 

 then we shall never know what the calicles are like in a specimen which escapes the ravages 

 of the sponge. 



The only other known Gonioporce from North-West Australia are too widely distributed 

 to be of any assistance in this respect, the nearest occurring over 200 miles away. 



a. Zool. Dept. 92. 12. 1. 211. 



. SUPPLEMENT TO GROUP IV. (Vol. IV. p. 85.) 

 Indian Ocean Forms. 



164. Goniopora Christmas Island 1. (G. Natalis prima.) (PI. VIII. fig. 7b.) 

 [Flying-fish Cove, coll. Andrews ; British Museum.] 



Description. — The corallum is explanate, thickening irregularly into convexities, with 

 lobate edges, pendent or encrusting, here sharp with projecting epitheca, there blunt and 

 rounded. The thicknesses are variable up to 1 cm. 



Tlie calicles are small, average 2 mm. in diameter, fairly deep, subcircular, but more 

 sharply polygonal when the skeleton is cleaned of animal matter. The walls are an irregular 

 lattice-work with jagged edges, here thin and ragged and there thicker, with traces of close, 

 narrow septal plates joined by a zigzag thread. The septa form narrow striae down the faces 

 of the walls, very variable in size, many quite rudimentary, some six or seven running out, bent 

 and irregular, to unite with a straggling, loose, reticular columellar tangle. The radial 

 symmetry Ls thus greatly obscured, only feebly shown in the septa, not found at all in the 

 columella or the pall, and only faintly visible in the interseptal loculi, wliich, owing to the 

 irregularity of the septa seldom seem to run deep into the coral. 



The section sliows a very irregular reticulum, the elements of which are coarse and 



