CORALS FROM THE CORAL RAG. 85 



the name of Thecosmilia trilohata} is one of the varieties. Sometimes the young coralhtes, 

 produced by a simple parent polyp, instead of forming a fascicular group, arrange them- 

 selves so as to constitute a short row, and do not separate immediately from each other ; it 

 may even happen that a few of these small series of corallites remain in contact laterally, 

 and thus assume the form of S^mjjJif/Uia. But these variations in the general form are 

 only met with in young specimens, and have never been met with in the older, large 

 Corals. 



The epitheca is well developed, and extends from the basis of the corallum almost to 

 the edge of the calices, but the septa are exsert. Sometimes this coating continues to 

 envelop two neighbouring corallites after these have become quite distinct internally, and 

 it presents numerous strong circular wrinkles or folds, which are closely set and very 

 unequally developed. When the epitheca has been in part, or totally, worn away, as is 

 often the case, the costa) or outer edge of the septa become visible, and appear delicately 

 denticulated, not very closely set, and alternatively somewhat more or less thick. Tlicre 

 does not appear to be any true walls, and the spaces situated between the costo-seplal 

 lamina3 are occupied by dissepiments. 



The calices are seldom circular, (as in fig. 1, Tab. XIII ;) they usually become very soon 

 oval, subtriangular, or lobated, and it often happens that two fossulae become perfectly 

 distinct some time before any corresponding change takes place in the margin, and are 

 united by common septa. The fossulag are small and rather shallow ; there is no appearance 

 of a columella, and the septa meet in the centre of the visceral chamber at a very short 

 distance from the surface of the calice. 



The number of the septa is extremely variable, and differs most especially according 

 as the calice belongs to a newly-formed corallite, or is more or less ready to multiply by 

 a fissiparous development. Similar modifications are always met with in fissiparous corals, 

 and renders it very difficult to come at the knowledge of the real specific characters of the 

 septal apparatus. But as far as that can be made out by the examination of the most 

 perfectly circular calices which must be supposed to belong to individuals that have not 

 begun to multiply in this way, it appears that the normal number of cycla is five ; the last 

 cyclum being more or less imperfect. The septa are thin, closely set, straight or slightly 

 flexuous, exsert, and terminated by an oblique arched edge, which is armed with delicate, 

 nearly equally develo])ed, denticulations. Those of the first three cycla are almost of the 

 same size ; those of the fourth cyclum not as thick towards their inner edge, and those of 

 the fifth cyclum are very thin ; all present on their sides slight granulations, arranged in 

 radiate series. 



This fine coral often forms large arborescent masses, one or two feet in height. The 

 specimen figured in PL 13 is eight inches high, and Mr. Charlesworth showed us in the 

 Museum of York a specimen, which, although incomplete, was more than one foot and a half 



' Milne Edvv. and J. Ilaime, Ann. Sc. Nat., s. iii, vol. x, p. 272. 



