of Heliophyllum and Crepidophyllum. oD 
which are strongly arched and have their convexities directed 
upwards. Though most largely developed in the external 
parts of the coral, the vesicles formed by the dissepiments of 
this series are variable in amount, and can hardly be said to 
constitute a distinct exterior vesicular zone, such as is so 
characteristic of the true Cyathophylla. 
With these remarkable points of agreement we find the 
following equally remarkable points of divergence, by which 
Crepidophyllum is distinguished not only from Heliophyllum, 
but from all other known genera of the Rugose Corals :—(1) 
The central tabulate area, in most respects, closely resembles 
that of Heliophyllum, the tabule being remote, often more or 
less arched, and sometimes uniting with one another. ‘The 
central portion of this area, however, is. shut off from the rest 
of the visceral chamber by a secondary investment or accessory 
wall, so that there is constituted a kind of central pipe or 
tube (fig. F), which is crossed by the tabula, and runs 
down the centre of the corallum. (2) The central tabulate 
tube thus constituted, however, is only rarely quite complete : 
usually it is open on one side, and its investment or wall 
becomes continuous at this opening with two of the primary 
septa, which run to the margin of the corallum. (3) By means 
of these two primary septa and the secondary wall there is 
thus enclosed a large, somewhat horseshoe-shaped septal fos- 
sula (fig. C), within which are contained two or, more com- 
monly, three short septa. (4) The remaining primary septa 
are continued inwards till they meet the wall of the central 
tube, with which they become coalescent. They do not, how- 
ever, extend into the interior of the tube; and there is there- 
fore no similarity between their arrangement and that which 
obtains in Heliophyllum, where a certain number of the pri- 
mary septa pass inwards to the centre of the visceral chamber, 
and become loosely connected with one another there. Indeed 
I am not acquainted with any genus in which any close 
approximation to the peculiar structure of the central portion 
of the corallum in Crepidophyllum can be found. ‘There is no 
other recorded genus in which the median portion of the 
central tabulate area is partitioned off by a distinct wall, with 
which all the primary septa are connected directly, and in 
which they terminate. 
The genus Orepidophyllum contains two species of corals 
from the Hamilton formation (Devonian) of North America. 
One of these corresponds with a portion of the group of forms 
which I formerly described under the name of Heliophyllum 
subcespitosum (Geol. Mag. new ser. dec. u. vol. 1. p. 58, 
pl. iv. fig. 9) ; and as it comprises the most typical members 
