THECID^E AND HELIOPORID.E. 239 



been placed in the plane traversed by the section. The septa 

 appear simply as blunt, triangular, irregular lateral ridges, with 

 ill-defined margins, having deep sulci at their bases, but always 

 leaving a considerable central area into which they do not pene- 

 trate. The interstitial tissue presents different appearances in 

 different parts, or under different conditions of preservation. 

 Sometimes — though by no means always — the interstitial tissue 

 can be recognised as divided into a number of small polygonal 

 areas by a network of dark lines, which represent the primitive 

 walls of the interstitial tubuli. Often these walls can only be 

 recognised by the slightly thickened nodes at their angles of 

 junction, or they may not be capable of detection at all. In the 

 same way, the minute central cavities of the interstitial tubuli 

 may appear in cross-sections as so many minute, irregular, 

 sometimes stelliform apertures, or they may be filled with 

 granular sclerenchyma, when the interstitial tissue appears 

 to be wholly solid. 



In longitudinal sections (PI. XI., figs. 2 c, 2 d), the visceral 

 cavities of the larger corallites are distinctly marked out, as 

 before, without any clearly-defined wall, and are seen to be 

 intersected by well - developed complete tabulae, which are 

 slightly flexuous, and sometimes anastomose to some extent 

 with each other ; while occasionally the section exhibits a 

 portion or the whole of one of the horizontal canals placing 

 contiguous tubes in communication. In sections of this nature, 

 the composition of the apparently dense interstitial tissue out 

 of numerous closely-approximated vertical tubuli can always 

 be recognised without the slightest difficulty. Moreover, these 

 tubuli, In sections of sufficient thinness, can always be recog- 

 nised as being bounded by perfectly distinct, apparently double 

 walls ; but their cavities appear to be more or less extensively 

 obliterated by a deposit of granular, not laminated sclerenchyma, 

 which seems, however, to be very irregularly distributed. No 

 tabulae can be detected, though it is possible that such struc- 

 tures really existed. 



From a consideration of the above characters, it will be 



