MR. H. 1ST. MOSELEY Oft THE STRUCTURE OF THE STYLASTERIDJE. 
i 
Around the mouth of the gastropore is a circlet of from about ten to fourteen 
dactylopores, arranged symmetrically at equal distances from one another and from 
the centre of the mouth of the gastropore. The mouths of these pores are elongated 
towards the axis of the gastropore, so as to open into, and become continuous with, 
the cavity of this latter pore (Plate 35, D Z, D Z). The openings of the dactylo- 
pores are continued down, as wide slits, for some distance on the upper part of the 
wall of the gastropore, so that the pores have, as it were, two mouths placed at right 
angles to one another and confluent with one another, the one opening to the exterior, 
the other into the cavity of the gastropore. The cavity of each dactylopore consists 
of a wide upper chamber in the region of the widely open mouth (Plate 35, T Z, T Z), 
and a narrow tubular continuation of this, which traverses the corallum in a direction 
parallel with that of the axis of the gastropore for about half the length of the 
latter. Against the outer wall of the pore is a small ridge-like excrescence, with 
an hirsute surface, which is the style of the dactylozooid (Plate 35, fig. 3, S'), and 
which is described by Pourtales as “ a rudimentary septum in the shape of a hairy 
fringe” (Pourtales, l.c., p. 34). 
The dactylopores in each cyclo-system are separated from one another by thin plates 
of calcareous matter, which are directed inwards radially towards the axis of the 
gastropore (Plate 35, fig. 3, P), and which at first sight have all the appearance of the 
septa of hexactinian corals, and have hitherto been mistaken for such by observers. 
They are, however, composed each of two thin laminae of dense calcareous matter, 
united by somewhat less compact calcareous substance, which is freely perforated by 
canals for the passage of offsets of the coenosarcal meshwork. The thin laminae are 
merely the juxtaposed walls of the adjacent dactylopores. These radially disposed 
plates, which may be termed pseudosepta, have their inner edges continued down 
the wall of the gastropore for a short distance beyond the margins of the mouths of 
the dactylopores, as well-marked vertical ridges, which soon become merged in the 
general surface in their course (Plate 35, fig. 3). 
The cylindrical masses formed by each cyclo-system are sometimes flat, often gently 
rounded at the top. Their summits are irregularly circular in outline, but have an 
indented border, the indentations corresponding with the centres of the pseudosepta 
in position, and representing the intervals between the opposed dactylopore walls, 
which are here not obliterated by growth of ccenenchym. 
The cyclo-svstems, when viewed from above in a line looking directly into the 
mouths of the pores, show, in all essential particulars, the same structure as that 
which occurs in Allopora 'profunda, which is represented diagram matically in Plate 35, 
fig. L3. The styles of the tenticular zooids, S S, appear as small projections in the 
interspaces between the pseudosepta, and were taken by Pourtales and others for 
septa of a second order. 
The cyclo-systems have been described as circular in outline of summit, because 
this may be regarded as their normal condition ; but very many of them are distorted 
