ent 12mm. : límm. The outer surface is rugged and ir-'| 
regular. Outer wall; the outer layer of the cup is occu- 
pied by a trabecular and vesicular tissue, and this appears 
to extend as a spongy network to the outer surface. The 
numerous septa, a little over 1mm. apart, are straight 
and clearly defined near the inner wall, and anastomose 
elsewhere to form a trabecular mesh; both the septa and 
the trabecular elements are united by thin curved vesicu- 
lar (dissepimental) laminae; the septa are pierced by 
numerous large pores, about .5mm. across, one pore to the 
mm.; the middie of the intervallum is varied in structure, 
at one point septa may pass without interruption from 
inner wall to near outer wall, at another trabecular tissue 
may intervene, giving the following succession—inner 
wall, septa, trabecular tissue, septa, trabecular tissue, 
cuter wall. The inner wall is a regular well defined layer 
with two rows of fairly large polygonal pores to each in- | 
tercept. The base, 20mm. across, is a mass of trabecular 
and vesicular tissue, without central cavity or well-defined 
septa, from which proceed root masses of similar tissue; 
the position of the inner wall is slightly indicated. Only 
one spécimen has been found, but it is a large, handsome 
end remarkable form, and well defined. We have named 
it after Professor T. Griffith Taylor. 
The genus may be defined as Metacyathidae in which a 
combination of septal and trabecular masses extends to 
the upper part of the cup, with strong development of 
vesicular (dissepimental) tissue. 
Metacyathus irregularis (new species). Fig. 29. 
A large irregular cone, with wavy outer wall, the inner 
wall following a similar contour; wide cavity; intervallum 
coefficient 10mm. : 30mm. The outer wall has an under- 
lying layer of fairly large irregular pores, usually two 
rows to each intercept, which are masked at the surface 
by a 
The septa are straight and delicate, a little over Imm. 
apart, and there is a great development of fine dissepi- 
mental tissue which often interrupts and confuses the 
septa; the septa are easy to expose in. transverse and 
tangential sections, but extremely difficult to expose in 
radial section; a small area darkened by manganese ob- 
securely indicates large pores (fig. 29c). The specimen is 
obscured by secondary silicification and по satisfactory 
trace of the inner wall was exposed; its place is taken in 
places by a thin structureless membrane (? secondary 
silicification). A length of 90mm. is preserved but the 
base is missing; we expect that this, when found, will 
agree with other members of the family. Two large root- 
like masses of vesicular and trabecular tissue occur in the 
lower part of the fragment. Only one specimen has been 
found. 
Taylors species Archaeocyathus dissepimentalis, 
though not identical with any of our species, has points of 
resemblance to Metaldetes conicus and Metacyathus irre- 
gularis, and may possibly be the upper part of a member 
of the Metacyathidae. 
Metacoscinus retesoptatus (new genus, new species). 
Fig. 27. i : 
The species presents a very remarkable combination of 
features. It is a conical cup; the upper part has numer- 
ous straight and regular septa, about Imm. apart; inter- 
yallum coefficient 5mm. : 15mm.; the external surface is 
waved by annular swellings, the inner wall remaining un- 
affected. The outer wall is a delicate network of irregu- 
lar polygonal pores, two or three to 1mm. The septa are 
an open net of large polygonal pores, irregular in size; 
they are perfectly straight and regular, without synapti- 
culae or dissepiments. "Tabulae of an unusual character 
are present in the lower part of the cup, but, if present in. 
the upper part, they must be at remote intervals; the 
tabulae are pierced by numerous small circular pores be- 
tween which are numerous round as well as irregular 
tubercles, proceeding from the upper and lower surfaces 
of the tabulae; many of these tubercles arch апа unite 
and become continuous with the septal network. 'The 
lower part of the cup at a diameter of 7mm. loses its cen- 
tral cavity; the septa become irregular; and the . whole 
area becomes filled with a mass of vesicular tissue; ex- 
crescences of the nature of buds or roots are present. The 
inner wall, in the upper regular part of the cup, is a deli- 
cate open network with large pores, two rows to each 
intercept. 
Severals pecimens have been found. 'The upper part 
strongly resemble's Taylor's Archaeocyathus retesentu. 
but we have not felt justified in claiming identity without 
Professor Taylor's sanction and direct comparison of the 
type specimens. The probable relationship is indicated hy 
our specific name “reteseptatus”; and, should the identity 
be admitted, the name would be Metacoscinus retesepta 
(Taylor). 
The genus may be defined as Metacyathidae with straight 
net-like septa in the upper part, and with tabulae present. 
Family HETEROCYATHIDAE (new family). Figs. 32, 34. 
The family is founded on two remarkable fragments 
which present some points of semblance to Archaeocya- 
| thinae, but in which a new i 
finer network and an outgrowth of fine processes. |і IO НИ emite, ЧУРА, RE 
growth of the walls from radiating centres. In опе of 
these specimens, fig. 32, two walls are present, with an 
| intervallum distance of 5mm., the walls being connected 
by a few coarse and irregular scattered rods or spicules. 
The specimen is part of a cone or cylinder some 50mm. 
diameter, fig. 22b and the lower part of fig. 32a герге- 
senting the outer wall, and fig. 32c and the upper part of 
fig. 32a representing the inner wall. Both walls show a 
series of straight or curved lines of tubercles radiating 
from centres some 6mm. apart. 'The tubercles of the 
outer wall are larger and longer than those of the inner 
wall, and they show a tendency to incline from the cen- 
tres of radiation and to anastomose. The preservation 
‚ and silicification of this specimen may be somewhat ano- 
malous, as the limestone matrix has indications of differ- 
entiation shown by the lines of shading in fig. 32a; it is 
therefore possible that the tubercles may represent casts 
in secondary silica of an originally porous structure; such 
casts are not uncommon in Archaeocyathinae from the 
locality. 
The porous structure of the other specimen, fig. 34, is 
quite definite; only a single wall, practically flat, is pre- 
served; the state of preservation and silicification is of 
the normal type. The pores, spaced on an average of two 
to 1mm., are in radiating lines separated by ridges which 
are most prominent near the centres of radiation; these 
centres are about 15mm. apart. There are a few knobs 
on the exposed surface which might possibly represent 
the bases of bars connecting this to a second. wall. Al- 
though the evidence is far from conclusive, we are inclin- 
