CORALS FROM THE PERSIAK GULF. 1031 



occurrence of a recent form as evidence of the close relationship 

 between the living and fossil fauna of that countrj^. He made 

 no attempt to discuss the significance of these forms in con- 

 nection with the existing classification of the Madreporaria, but 

 dismissed the subject with the single statement — "the broad 

 distinction usually made between perforate and non-perforate 

 corals breaks down." 



The occurrence of another recent species of this problem- 

 atical genus again forces the subject on the attention of the 

 systematist. That it is a Trematotrochus is, I think, unquestion- 

 able. It agrees with Dennant's latest definition in all points 

 but one : — He ha,s described the septa as being solid ; but in 

 T. zelandice the septa are incomplete, large fenestrations occurring 

 at the point at which the tei'tiaries join the secondaries, and the 

 primaries and secondaries join the false columella. Realizing 

 that the existence of these fenesti'ations and of the pores in the 

 outer wall might again be doubted (having already admitted 

 that theso latter are minute and difficult to see), I have ground 

 sections of one specimen, thinking that actual photographs would 

 be more convincing than external observations. These sections 

 have been made at intervals of -^ mm. and the entire coral has 

 been ground away; every section was photographed at a magni 

 fication of 20, and from these photogmphs a wax model has 

 been constructed. There is, therefore, a permanent record of the 

 structure of this most interesting coral, and it would be idle to 

 deny the existence of either thecal pores or septal fenestrations. 

 But in spite of this, I think that it is rightly placed among the 

 Imperforate corals. 



In text-fig. 218, A and C are diagrammatic representations 

 of transverse sections of portions of the wall of a Heterocyathus 

 and a Heteropsammia (after Bourne [6 ■), for comparison with B, 

 a transverse section of T. zelandio}. The Heterocyathus has a 

 solid theca embedded in stereoplasm. The Heteropsammia has 

 a porous spongy wall. A glance at the Tre-nmtotrochus will 

 show that the extremely thin wall is incomplete, and that here 

 and there are small communications between the interseptal 

 chambers and the exterior ; but it could not be regarded as a 

 '-porous coenenchyme" such as is diagnostic of the division 

 Perforata. The condition is much more that of a Fungiid which 

 lias grown conically instead of spreading out with a widely open 

 oral surface, with that part of the wall between the j)erforations 

 representing synapticula. In some of the fossil forms the per- 

 forations are so large and the intervening solid structure little 

 more than a narrow bar, that the comparison with synapticula 

 is more striking than in the present instance. The analogy, 

 however, must not be pressed too far, but it widens the o-ulf 

 between Trematotrochus and the Perforata, and tends to justify 

 its inclusion among the Imperforata. 



Another point to be emphasized is the localities in which the 

 species are found. Hitherto the eight species of the genus were 



70* 



