54 R. G. Garruthers — The Garboniferous Corals 



and description amply suffice, there being no risk of confusion with 

 any other Tournai fossil. 



Tahdce. — Apparently it has always been supposed that this beautiful 

 little coral had no tabulae, being open from top to bottom. Chiefly 

 on this ground the genus has been placed in a family apart from 

 other Palaeozoic corals (the Cyathaxonidae of Milne-Edwards & Haime), 

 which has supplied material for much ingenious speculation as to 

 possible descendants amongst the Hexacoralla. As a matter of fact, 

 tabulae are present in abundance ; they are, however, very thin, and 

 doubtless the misconception arose because these little fossils, as found 

 at Tournai, are invariably silicified, and so fragile that tabulae are 

 not apparent on breaking open a specimen. But if the coral be first 

 steeped in hot Canada balsam, and then carefully ground down, a thin 

 section can be made, in which the tabulae are clearly seen, rising 

 upwards to the strong central columella (PI. Ill, Fig. 6). When the 

 specimens are calcareous, as are most of those from British and Irish 

 localities, vertical sections showing tabulae are readily obtained, 

 although microscopical examination is generally requisite, owing to 

 the very small size of this species. 



Columella. — At the extreme tip of the corallum no trace of 

 a columella can be seen, the septa simply meeting at the centre.^ 

 When eight or nine septa have developed (the number varies con- 

 siderably) the columella appears, and quickly attains prominence. 



Both vertical and transverse sections show that the columella is, 

 structurally, quite independent of the septa. Thin transverse sections 

 (PI. ,111, Fig. 4) show the calcite fibres radiating outwards from 

 a central point, usually dark ; concentric coloured growth-rings are 

 usually to be seen round this axis. The major septa are fused 

 against the columella, but in no way partake in its construction, for 

 their crystalline fibres and central dark or white line terminate as 

 soon as the columella is reached (Fig. 4, PL III, is too small to 

 illustrate this point, which is better seen in the section of C. rushiana, 

 Fig. 10). 



Vertical sections also demonstrate the independence of the columella 

 and septa. The section shown in Fig. 5, PI. Ill, is cut down the 

 axial plane of the coral, along the centre of both the columella and 

 of the major septum on each side. Here, again, the arrangement of 

 the ciystalline fibres (at right angles to which growth takes place) 

 indicates the absolute lack of continuity between septa and columella, 

 although these latter are in complete contact. 



Compared with Loplio'phylhim, the columella of a Cyathaxonia may 

 be regarded as an extreme case of the upward bending of a septum in 

 the centre of a coral, developed to such a degree that severance has 

 taken place and an independent structure arisen. This conception is 

 here illustrated in diagram form, figures of two other forms being 

 added to show progressive divergence from a typical Zaphrentis, 

 although a phylogenetic connexion is by no means implied. 



^ The primary septation of Cyathaxonia cormi forms the basis of a paper by 

 M. Faurot (" Affinites des Tetracoralliaires et des Hexacoralliaires " : Annales 

 de PaUontologie, torn, iv, 1909). This important communication deserves 

 separate treatment, and is therefore only referred to here. 



