68 MADREPORARIA. 
ridges are thin, wavy, and irregular. The septa begin very near the top of the wall-ridge, 
as small irregular projections often expanding into T-shaped flakes. Deeper down, these flakes 
are More prominent, meet and fuse, and together form an inner synapticular wall, which forms 
the true inner rounded skeleton of the calicle within the larger irregular polygonal areas 
circumscribed by the wall-ridges. This inner wall is not here very symmetrical or regular, and 
there is a tendency for it to melt down with the wall-ridge into a reticulum. 
The septa are very thin, but are conspicuous owing to the frosted granules and the large 
plate-like pali, which show that the septal formula is typical. The pali form together a large 
ring, which at first sight appears very irregular, and to consist of a far greater number than is 
typical for Porites; and, further, being flattened, they are seen not to be radially symmetrical- 
These appearances are due to the fact that the large pali of the primaries slope towards but do 
not fuse with the smaller pali of the secondaries, hence each principal palus is represented by 
its two elements not fused together. It is a V-shaped body, but with the points not touching. 
The fossa is large and open, with a very thin central tubercle. Deep down beneath this the 
columellar tangle is circular and compact. 
Round the projecting sides of the stock the skeleton tends to become an open flaky 
reticulum, indicating that this part of the colony was growing rapidly. On the under surface 
(PL. V. fig. 8) the T-shaped flakes which thicken the wall-ridge become very pronounced, and 
form a shelf all round the calicle; while the septa become more conspicuous and echinulate, 
the plate-like pali are resolved into so many distinct echinulate granules, and are thus less 
conspicuous. 
The section is strongly trabecular, the individual trabecule varying greatly in thickness. 
This flat-topped growth-form, already described in the Porites of the Fiji Islands (see 
Nos. 4, 7, and 22), here occurs again in this Ellice Island group (see, for example, Nos. 6, 9); 
it is so far unknown elsewhere (see Introduction, p. 22). 
The coral is peculiar in the arrangement of its pali. As a rule, when in large deep 
calicles the septa slope and only fuse deep down (see Introduction, pp. 18, 20), the pali— 
which at times seem to owe their existence to the fusions of the septa—are not developed. 
Here we have septa with very pronounced pali sloping towards one another but only fusing 
deeper down. The fact that the primary pali are so much larger than the secondary is of 
importance ; in Goniopora it sometimes appeared as if the pali were associated with the 
secondaries rather than with the primaries. : 
Mr. Gardiner has called attention to the close similarity between P. Fiji Islands 4 and 
this coral by uniting them under one specific name, “¢trimwrata.” This name, as described in 
the Introduction, p. 16, I have applied to a method of wall-thickening which is common in 
Porites. The calicles of the two forms are built on the same essential plan, but are larger in 
the Fiji form, and the rings of pali are not so conspicuous. We have no means yet of knowing 
whether the likeness is due to close genetic affinity, or to parallel developments under the 
influence of environments which closely resemble one another. 
Fig. 6 is from the upper, fig. 7 from the lower, surface. 
a. A large fragment from an edge of a flat-topped growth. Zool. Dept. 1904. 10.17. 26." 
b. Zool. Dept. 1905. 1. 19. 16. 
