364 MESSES. H. A. NICHOLSON AND E. ETHEEIDGE 



great regularity ; but the raost remarkable feature is the presence 

 in many of the corallites of a longitudinal or vertical ridge, par- 

 tially dividing the visceral chamber into two portions. Sometimes 

 this ridge is barely perceptible, or even wholly absent ; at other 

 times it extends halfway across the visceral cavity ; and some- 

 times it nearly touches the opposite wall. In some parts of a 

 large corallum almost every corallite may possess one of these 

 ridges, whilst in other parts not a single structure of the kind is 

 to be detected. Nor is there always but one of these ridges 

 within the cavity of a single corallite, but there may be two or 

 three, in which case one is generally much larger than the others. 

 The significance and nature of these vertical ridges will be con- 

 sidered immediately ; but it is in the meanwhile necessary to 

 briefly glance at two corals which are very closely allied to ^. sep- 

 tosa, and which undoubtedly belong to the same group. 



One of the corals in question has been found by us in consi- 

 derable abundance in the Carboniferous Limestone of the north 

 of England. It occurs in the form of hemispkerical, subglobose, 

 or irregular masses, with a convex surface, often of considerable 

 size, composed of prismatic corallites, which may or may not be 

 arranged in successive superimposed layers half an inch or more 

 in thickness. In other instances, the corallites are tall and basal- 

 tiform ; but in either case they are " erect," in the sense in which 

 this term has been previously used. The calices, therefore, are 

 not ohlique, but are like those of A. septosa. There are gener- 

 ally about five calices in one line ; and though irregular in shape 

 and size, they are essentially subhexagonal in shape. The walls 

 of the corallites are thick ; but we have wholly failed to discover 

 any mural pores. In transverse sections, certain of the corallites 

 show a strong tooth-like process on one side, sometimes opposed 

 to two smaller teeth ; and these ridges may be short, or they may 

 extend halfway or more across the visceral chamber, whilst tliey 

 are wanting altogether in a large number of the corallites. In 

 longitudinal sections the corallites are seen to be crossed by re- 

 mote horizontal tabulae, which are not uncommonly placed at the 

 same level in contiguous tubes, and which may be wanting alto- 

 gether in certain layers. 



The corals we have just described present a very close resem- 

 blance to A. septosa, Elem., from which they can only be distin- 

 guished by the smaller size of the corallites, the somewhat thicker 

 walls of the tubes, and the slightly more remote tabulae. Whether 



