TRIBE II. CAR YOPII YLLACEA. 353 



wholly wanting, and the texture of the corallum, in a transverse sec- 

 tion, is simply porous or spumous. It is quite probable that this 

 absence of distinct lamellae, and the cellular texture instead, may 

 depend, as in the Porites and Gonioporae, on the polyps being long 

 exsert when expanded, and only their lower portions, below the visce- 

 ral cavity, secreting lime. 



In a vertical section, the transverse septa are sometimes seen to 

 extend quite across the whole interior, while in other species, they are 

 confined to the middle portion, or become almost obsolete. They are 

 seldom regular in their intervals, or in a single range ; on the con- 

 trary, there is commonly a confusion of transverse cellules through 

 the centre of the corallum. In several species, the septa, instead of 

 crossing transversely, are oblique, and merely overlap about the medial 

 line; while in others, they extend upwards very obliquely, from either 

 side, and meet more or less perfectly, in an axis to the corallum. Some- 

 thing apparently analogous to this is seen in certain Astraeas : the 

 lamellae instead of being finished out entire, have the inner edge pecti- 

 nated obliquely upward, the pectinations extending towards a medial 

 line. Yet the Cyathophyllidae differ essentially in having, for these 

 oblique lines, prolongations of the intermediate dissepiments which 

 unite the lamellae, and not of the lamellae themselves. It appears to 

 be a general principle that, while these intermediate dissepiments are 

 altogether subordinate to the lamellae in the Astraeidae, in the family 

 under consideration, they are usually stouter than the lamellae; the 

 transverse septa alluded to, result from their continuation, and more 

 or less perfect union, through the central portions of the corallum, 

 where the radiating lamellae are often wanting. The axis, in species 

 with oblique ascending septa, is sometimes formed by a convolution 

 of the septa or plates, as in some Columnarise. In the Sarcinulee, the 

 septa extend quite from the sides, and appear like a series of funnels 

 inverted upon one another, with a solid axis along the central line. 



Michelin first pointed out that certain species have on one side of 

 the cell a narrow triangular depression or cavity, with one or more 

 of the lamellae in part wanting, and designated the group Caninia. 

 This structure cannot be considered altogether anomalous, when we 

 consider that some Madrepores and Zoanthidae have one tentacle dif- 

 ferent in size or colour from the others; and we need not look for an 

 explanation of it, to an analogy with the siphuncle of the Ammonite. 



We observe a farther difference between the coralla of the Astraees 



89 



