16 
to a Favosites, and the other to a Heliolites, but in both cases the area of 
attachment is restricted by the form of the supporting coral. In other species 
(e.g., II. crcitus ) there is every appearance of a much expanded base, as if 
the colonies had adhered to a large object, neither is this surface in one plane, 
but undulating and uneven. Nicholson says the colony of Ilaly sites in the 
early condition is not unlike that of an Aulopora. 1 
4. Epitheca. — The epitlieca, notwithstanding the change in mineral 
composition, is usually well displayed on the exterior of the corallites as fine, 
closely set, fluctuating, or rolling, transverse lines, continuous around all sides 
of the fenestrules. As Nicholson has observed, 2 it does not take any part in 
the formation of the transverse walls bounding the larger corallites. From 
Mr. Etheridge’s remarks on his II. catenulcirius, var. Fieldeni from the Arctic 
Region, already referred to, it would appear as if a basal epitheca existed also. 
5. Fenestrules. — These are the “masse lacuneuse ” and “irregular 
reticulations ” of Milne-Edwards and Haime, and the “ interspaces ” of Lambe. 
The fenestrules are enclosed by the union of the corallite chains, and vary 
considerably in size within certain limits ; from two millimetres in both 
diameters up to fourteen millimetres by twelve, twenty-three by one, and 
twenty-five by three. The outline is subject to much variation, they may be 
round, oval, polygonal, labyrinthine, convoluted, irregularly oblong, or almost 
square, even S — and dumb-bell shaped ; the polygons, usually longer in one 
diameter than the other, may be pentagonal, hexagonal, or heptagonal. The 
walls, which are equally the coalesced corallite walls, are always strong, and 
either fiat and barely ribbed longitudinally, or transversely undulate in 
varying degrees, owing to the more or less farcimentiferous outline of the 
corallite chains. On the whole, the largest fenestrules appear to be in 
U. australis. There is not any particular arrangement of these interspaces 
except in one species (R. psristephesicus), where there appears to be an effort 
to cluster or revolve round certain centres. 
6. Corallite Chains. — I employ the term corallite chain to distinguish 
the zooidal apertures arranged in linear sequence along each angle of a 
fenestrule, and by ultimate intersection enclosing the latter. They arc to 
some extent the “ grandes lames flabelli formes ” of Milne-Edwards and Haime, 
and the “ lamina” of Nicholson and Lambe. The outline of a chain depends 
1 Nicholson, Man. Paleontology, 3rd Edit., 1S89, p. 339. 
2 Nicholson, Tah. Corals Pal. Period, 1879, p. 227. 
