iS) 
CO 
nN 
tooth of the edge bears on it a number of striae, sometimes 
veiled superficially by the granulations, but always apparent 
after polishing the surface. This simply means that the septal 
fold in Mussa is thrown into a few very wide, deep pleats, corre- 
sponding with the broad ridges of the septal surface, and then 
that each wide pleat falls into a system of smaller, pitted pleats. 
Transverse and longitudinal sections show a departure from the 
d.n 
Fic. 4.—Transver-e section of part of septum and wall, showing the radiating lims of 
) I points,’ and also the wavy cross-lies 
corresponding to successive ‘‘ growth-lamella” (high power) ; gran. = granulaticn. 
Structure passing out from the ‘dark 
strict bisymmetry of fibres and uniform distribution of axes 
observed in the striated septum. 
The septum of Heliastraea presents a further variation of the 
Mussa scheme. In microscopic transverse sections (Fig. 5), the 
fascicles are especially closely grouped in the middle, bulging 
part of a ridge, and are there arranged almost circularly ; 
while, at the narrower ends of the ridge, the fascicles are 
further apart and are set bisymmetrically. 
Next, in some Astreeid types, and conspicuously in the family 
of Fungidze, similar large ridges are present, but they are 
elliptical, squarish, or roundish in shape, indicating a still 
closer bundling together of the fascicles. As a fact, the fascicles 
radiate out around a common ridge-axis, although the indi- 
viduality of the fascicles is still maintained. This is proved 
1G. 5.—Thick central portion of a septal ridge of Heliastraea formed bya 
radio-symmetric trabecula (77.): and the thinner lateral wings of the 
same ridge passing into corresponding furrows—only bi-symmetric 
trabecule (¢».) are present in these (magnified 70 times). 
by the distinctness of emergent ends of fascicles in the thinner 
septa, and also by the inherent structure of parallel-set striae. 
In thick septa, however, the emergent ends of fascicles coalesce 
to form the characteristic large granulations. These are the 
granulations, composite in structure, which give rise to ‘ pseudo- 
synapticulee”’ (Pratz), inasmuch as their large size enables those 
at the same horizon on adjacent septal surfaces to coalesce 
across the interseptal loculus. 
Thus I have passed from a bisymmetric to a radio-symmetric 
arrangement of fascicles in septa. Throughout all these septal 
types, certain definite laws of growth can be observed. Each 
NO. 1421, VOL. 55] 
NATURE 
[JANUARY 21, 1897 
row of fascicles (Fig. 6) added along the growing septal edge, 
indicates a single period of growth in the existence of the polyp. 
I have, therefore, applied the term of ‘‘ growth-segment ” to 
signify one such addition made to the height of the septum ; 
the term ‘trabecular part ” to a pair or group of fascicles ; and 
I define the term ‘‘ trabecula”’ asa series of ‘‘ trabecular parts ” 
laid down during successive growth-periods of the polyp. 
In the case of most Turbinolids the dasa/ 
deposit laid down during a growth-period is 
in the form of a few more lamella added, 
without any spacial interruption, to the 
lamellae of the previous growth-period. In 
the case of tall Astraids, Stylinids, &c., the 
basal deposit of each growth-period forms a 
new calycinal floor. And the space between 
any one calycinal floor and its predecessor 
corresponds to one growth-segment of the septum. 
This is a most important relation, and is one 
which I find also holds good for the synapti- 
culate form of base. The ‘‘ true synapticula” 
in Fungia and its allies would thus be homo- 
logous with a basal dissepiment, and is prob- 
ably an acquired feature in them, modified 
gran. from a more primitive dissepimental base. 
Certainly one finds that dissepiments dwindle 
or even disappear in these synapticulate types. 
The question arises of the possible advantage to 
the polyp of the synapticular in place of the 
dissepimental floor. The advantage is, at 
least, two-fold: (1) mechanical, as a support 
equally strong, although proportionally lighter ; 
(2) physiological, as a means of extending the 
internal ccelenteric space by the canalicular 
elongations, which are supported between 
synapticulee. 
Vhe septum of the Eupsammidz is, as in 
the Fungidz, frequently porous, and the 
interseptal loculi are bridged by the coal- 
escence of septal ends, by pseudosynapticulz, or by true 
synapticule. The pores of the Eupsammid septum, however, 
may occur not only between the adjacent ‘‘ trabecular parts” 
go. am. 
gr 
tr. p. 
Fic. 6.—Radial structure composed of a ~ costal” 
wing, 
ence’ between the two wings. The single trabecula marked “#7.” is 
composed of seven trabecular parts (¢7./.) representing seven successive 
growth-periods. The trabecular parts between any two growth-curves 
(g.c.) form a ‘septal growth-segment,” and represent one growth- 
period (diagrammatic). 
“septal” wing and a 
The trabecule diverge right and left from an “‘avea of diverg- 
of a growth-segment, but also between the successive ‘‘ trabe- 
cular parts” in any one trabecula. Hence it is possible for a 
‘trabecular part” to be absolutely free from its neighbours n 
