360 ON THE ANATOMY OF A SUPPOSED NEW SPECIES 



The septa naturally decrease in thickness from the primaries to the tertiaries. 

 All are relatively thin and little perforated, with almost smooth edges and sides 

 covered with low, blunt, somewhat distant granules, not arranged in any determinate 

 manner. Synapticula are absent. 



The columella closes in the axial fossa below, and owing to the narrow directive 

 septa, added to the slightly elongate shape of the calice, is very distinctly oval. It 

 varies in dejjth, in the larger corallites being situated from 4 to 5 mm. below the 

 edge of the calice. In the 3'oimgest separate corallite, that I have examined, there 

 appeared to be a true columella, arising from a basal plate, but in the older calices 

 it has a spongy appearance, and seems to have been formed principally by the ana- 

 stomosis of a large number of trabeculae from the septal edges. 



I have not attempted to examine the minute anatomy of the corallum in any 

 detail. The skeleton in all the specimens is everywhere much bored into by algal 

 filaments, which although found principally in the deeper lying parts extend in places 

 to within '2 mm. of its surface. 



SECTION II. 

 General Anatomy of the Polyps. 



Composition of the Colony. The corallum, as mentioned before, is everj-where, 

 except over its attached base, covered by the polyps. The latter may be regarded 

 each as an independent individual capable of leading an independent existence. All 

 the polyps however are connected together by the coenosarc, which consists of a 

 number of canals separated from one another by a double layer of endoderm with 

 the structureless membrane between. These canals run from poh^) to pol3-p — branching 

 perhaps at the bases of the projecting corallites — and serve to put the gastrovascular 

 cavities of the diflerent pol}'ps in free communication with one another. 



A similar arrangement is found in Pocillopora and all corals, so far as I am 

 aware, which have a well-developed peritheca, save that in some genera, as pointed 

 out by Fowler (10), the canals have been pushed apart from one another and the 

 external wall, consisting now of a double layer of ectoderm with the structureless 

 membrane between, lies directly on the corallum. In imperforate Madreporaria the 

 coenosarcal canals may be said to commence from the edge of the calice, while in 

 perforate forms, especially in such forms with partially free corallites as Coenopsammia, 

 no such sharp line of distinction can be drawn, since the iutracalicular portions of the 



