PROF. P. MARTIN DUNCAN ON MICRABACIA CORONULA. 563 
pase, although circular in outline, is not horizontal in any of my 
specimens, for there is a slight but decided hollowing which permits 
the coral, not to rest upon the whole surface of the disk-lke base, 
but only on its edge or circumference, 
At first sight this is an unimportant peculiarity, yet it clearly 
relates to the nature of the synapticular structures, and to the more 
or less perfect canal-system within the coral. Were the base per- 
fectly flat, there could be no ready communication between the 
outside medium and the interseptal spaces from below. Such a 
communication, although distinctly opposed to the idea of a Coelen- 
terate organism, is evident enough in many of the Fungine, and 
especially in the genera Yungia and Herpolitha. In the Fungi 
especially the base is hollowed out and the corals rest on the edges 
of the disk, leaving a space for water between the supporting sub- 
stance and the centre of the base. 
The coste are as described by MM. Milne-Edwards and Jules 
Haime; and it facilitates the comprehension of their relation to the 
septa by stating that each costa bifurcates at the circumference of 
the base of the coral, and that the right-hand bifurcate process of 
one costa unites with the left-hand bifurcate process of its neighbour, 
and that a septum is prolonged from this union (fig. 3). 
This arrangement necessitates the septa being continuous in 
direction with the intercostal spaces, and not with the coste. 
The regular perforations of the intercostal grooves are very sym- 
metrically placed, and add to the beauty of a well-preserved fossil. 
These openings lead up into the coral, between the septa into the 
interseptal loculi. 
‘These spaces are bounded at the sides by coste, and internally and 
externally by bars which are the lower ends of synapticula, and not 
portions of a theca or wall. 
In order to comprehend this statement and to examine the synap- 
ticula at the base and higher up within the interseptal loculi, it is 
necessary to fracture a coral, so as to separate some septa from those 
next to them. 
On examining the exposed sides of the septa it is found that 
these moderately close, unequal lamine are solid. They are den- 
ticulate at the free edge, rather thin here and there near the serrated 
margins, and more or less ornamented with lines of granules which 
never unite across the interseptal spaces with their fellows of the 
opposed septa. 
But there are fractured surfaces seen on the sides of the septa, 
situated on irregular eminences which vary in length, and have a 
decided trend from the base or lower part of the septa, radiating 
more or less, to about three-quarters of the distance from the lower 
part of the septa to their upper free edge (fig. 1). Sometimes the 
projections approach nearer the edge.. ‘The projections are synap- 
ticula broken across ; and it is noticed that whilst some are long and 
extend uninterruptedly over much of the septal side, others are 
shorter, and that a few bifurcate. The synapticula are stout, 
moderately thick from within outwards, and tolerably close one to 
