chief Generic Types of PaJeeozoic Corals. 427 



ventral side. Generally speaking, there is only a single fos- 

 snla ; but sometimes there are subordinate depressions, which 

 appear to be of the same nature, though these have not been 

 as yet sufficiently investigated. The most characteristic fea- 

 ture about the fossula, however, is the fact that it is formed 

 by the coalescence of a greater or less number of the septa, 

 which unite with one another towards the centre of the visceral 

 chamber and form the walls of the fossette (PI. XII. figs. 5 & 6). 



The mode of formation of the fossula is thus entirely dif- 

 ferent in Zaphrentis from what obtains in Amplexus. We are 

 not aware that this peculiarity has been noticed before ; but all 

 transverse sections exhibit it, and we believe it to be a cha- 

 racter of primary importance in the definition of the genus. 



The septa in Zaphrentis are differently developed in different 

 species — but are usually remarkably thick towards the margins 

 of the corallum, and become attenuated as they approach the 

 centre of the visceral chamber. Indeed in some forms the 

 septa are so much thickened towards the periphery, that they 

 coalesce at their outer margins, and form a thick false wall. 

 They also differ as to their extension inwards. Sometimes 

 they reach quite to the centre of the visceral chamber ; but at 

 other times they fall short of this, and leave a small central 

 area occupied solely by the tabulre. In silicified specimens, 

 however, the septa can often be traced on the upper surfaces 

 of the tabula? as far as the centre of the corallum (see PL XII. 

 fig. 6, Zapli. patula). Small secondary septa are usually de- 

 veloped, one between each pair of primary septa ; but they 

 do not appear to be universally present. According to Milne- 

 Edwards and Hairae, the free edges of the septa are denticu- 

 late, as they project into the calice. This peculiarity, how- 

 ever, so characteristic of the genus Ilcliophyllum, we have not 

 been able to recognize in any of the species of Zaphrentis 

 which have come under our notice. The development of the 

 interseptal dissepiments also seems to vary in different species 

 of the genus. In some of the Devonian species these struc- 

 tures do not appear to exist at all ; and in others they are 

 but sparsely produced. Delicate curved dissepiments, how- 

 ever, with the convexity upwards, are commonly developed 

 towards the circumference of the corallum ; but in no case, 

 are the dissepiments developed to such an extent as alto- 

 gether to obliterate the tabular or to form a distinct external 

 vesicular area. This latter structure is not found in any 

 true Zaphrentis-, and we shall hereafter show that the so- 

 called Z. cylindrica of the Carboniferous, in the structure of 

 which an outer zone of vesicular tissue is a conspicuous feature, 

 is truly referable to the genus Ci/athojiln/Unm. 



30* 



