238 TABULATE CORALS. 



chymal tubuli " of Heliolites and its allies, which we only know 

 by the quite recent researches of Moseley to be not truly of a 

 coenenchymal nature. When, in fact, we examine exceedingly 

 thin sections of Thecia, whether these be either transverse or 

 longitudinal, we find that the apparently dense and compact 

 tissue which seems to separate the large corallites is in truth 

 composed of vertical tubules, which we must now regard — 

 from what we know of the Helioporidiz — as being properly 

 small corallites, tenanted by a special set of zooids. So far, 

 then, the structure of Thecia is essentially identical with that 

 of Heliolites, but there are some very important differences to 

 be observed in these types respectively. In Heliolites and its 

 relatives, both the larger and the smaller corallites are enclosed 

 by thin but perfectly distinct walls, the former having well- 

 developed septa and remote tabulae, the latter having no septa 

 and crowded tabulae ; while there is no sufficient evidence of 

 any lateral communication between the visceral cavities of the 

 larger corallites. In Thecia, on the other hand, the large coral- 

 lites cannot be proved, even in the thinnest sections, to possess 

 any proper wall, their boundaries not being clearly separable 

 from the dense tubes of the interstitial tissue ; the septa which 

 they possess are not at all similar to the thin inflections of the 

 wall which constitute the "pseudo-septa" o{ Heliolites -dsiAxXs, 

 allies, but are thick and irregular ridges ; while the interstitial 

 tubuli — so far as can be certainly ascertained — are destitute of 

 tabulae. There is, moreover, the curious feature that the larger 

 corallites are placed in direct communication by means of dis- 

 tinct, often tortuous, horizontal channels, which traverse the 

 interstitial tubuli. 



If we examine a thin transverse section of Thecia Swinder- 

 iiana under the microscope (PI. XL, figs. 2 a, 2 b), we observe 

 that the visceral chambers of the large corallites are quite dis- 

 tinctly marked out, though nothing of the nature of a differen- 

 tiated bounding-wall can be detected. Here and there the 

 visceral cavities of two contiguous polypes may be seen to be 

 directly connected by a horizontal canal, which happens to have 



