158 DESCRIPTION OF CORALS. 
were apparently less so than in the French specimens ; it must also be stated 
that M. Michelin does not allude to the diverging mode of growth, or to 
studded laminz. The lobated or ramose processes could not alone be assumed 
as a specific distinction, analogous offsets sometimes occurring, sometimes not, 
in other corals, as Heliopora carulea; and it must be remembered that Ast. or 
Porites panicea was a polymorphous polypidom. (Iconog. Zoophytol. p. 160, 
Observ.) 
The stars irregularly distributed on the tabular surfaces (fig. 7) had a pre- 
vailing diameter of about half a line, but they varied considerably in characters. 
Frequently all signs of a definite boundary were wanting, the edges of the 
lamelle blending with plates of the reticulated structure, and not being con- 
nected by curved lamine as in persistent walls. These apparently growing 
stars were either on a level with the surrounding surface, or slightly raised 
above it (fig. 7a). In other instances the stellular cavities had a bold yet un- 
equal boundary, and the lamelle were united by arches ; while the terminal cup 
varied in depth, though its contents were not observed to rise above the 
margin. Surfaces, believed to exhibit different stages of development in the 
intermediate structure, consisted of thick points variable in shape and size, and 
occasionally blended into irregular vermiform plates of limited range. The 
intervals between the points or plates were very narrow, and generally occupied 
by matrix or iron pyrites. They had little extent, but within their limits they 
composed apparently solid layers, which conformed to the variable level of the 
specimen, and concealed all subjacent structures. The small points were some- 
times lineally arranged, but more often irregularly disposed. In side fractures 
or sections, these laminz were more or less exposed, ranging for short distances, 
and without any definite position ; but in one instance four successive layers 
were exhibited about a quarter of a line from each other (fig. 7 a), and the points 
were lengthened into filaments which extended from one lamina to the next. 
The tabular surfaces presented also occasiona!ly a perfect blending of structures, 
by which the stars were so commingled with the interspaces as to be imperfectly 
distinguished, and sometimes there was a general thickening of the exterior, due 
probably to abrasion. 
The best vertical section (fig. 76) exhibited fully the diverging mode of 
growth ; but the central line had apparently been occupied, to a considerable 
extent, by an extraneous body, and another similar cylindrical cavity existed 
elsewhere. The component structure was very confused, consisting chiefly of 
