512 Mr. H. M. Bernard on the 



the single specimen of a new species, which I propose to call 

 T. auricularis, some of the accessory lobes certainly spring 

 from the border-line between the living and the dead coral. 

 Even in some other cases, where the hindrance cannot be so 

 easily concluded, it is probable that it occurs. 



Fourth Type of Growth : Foliate. — I propose to group under 

 this heading all those cases in which the edge of the cup grows 

 up into wavy fronds more or less erect, which may fuse irregu- 

 larly together in every imaginable way. Tlie fronds may be 

 very deep and wide apart, or else very narrow, in which case 

 they are generally very closely packed. Some very remarkable 

 variations in their method of growth are found. In some the 

 under surfaces of the fronds {%. e. the surfaces without polyps) 

 are close together, while the spaces between the polyp- 

 bearing faces of the fronds are wide apart ; this is the arrange- 

 ment one would naturally expect to be most suitable for the 

 life of the polyps. There occur forms, however, in which the 

 polyp-bearing surfaces almost touch, and even fuse, while 

 the spaces between the under surfaces where there are no 

 polyps are wide apart. In all cases the early cup is soon 

 completely overgrown and obliterated. 



Fifth Type of Growth: Mesenteriform. — This name is 

 borrowed from Lamarck's species, T. mesenterina. The 

 growing margin creeps outwards more or less horizontally, or 

 even downwards. It is divided into lobes, which are sepa- 

 rated by folds bent vertically upwards (c/I diagrammatic 

 drawing, PL XIX. fig. 6). These folds are grown round, and 

 then form open cylinders or closed knobs or finger-shaped 

 processes. As the corallum expands the stalk is completely 

 obscured, and its origin from an early cup could never have 

 been guessed. The method of growth is, however, fairly 

 uniform, and is pronounced enough to be recognized without 

 difficulty. There is, further, fortunately a specimen in the 

 National Collection at about the stage figured in the diagram 

 (fig. 6). It forms a connecting-link between the early cup 

 and the flat, nodulated, encrusting masses which show the 

 " mesenteriform " method of growth only along their ex- 

 panding margins. 



Sixth Type of Groxoth: Tabulate. — I have adopted this 

 name to designate a curious method of growth by no means 

 infrequent. The cup evidently grows out rather flat, with 

 slightly curled-up edges. As it dies down a fresh layer, 

 appearing to start from the edge of the old, not only expands 

 further, but spreads back over the old, and not always in 

 contact with it, but arching over. An old stock thus shows 

 several tiers of more or less horizontal coralla, which may be 



