PROCEEDINGS OF SECTION C. 435 



As regards the general form, one sees turbinate specimens 

 Tesembling closely many of the Rugosa, such as Om.'phj'ma. But on 

 the other hand in some examples the rim of the calice spreads, as in 

 the flower of the convolvulus. This recalls no coral — ^with the exception, 

 perhaps, of Sanidophyllum — .in which the corallites are united by a 

 similar expansion ; not, however, containing a complete septal system, 

 as in the Archaeocyathinae. The restoration of the hexactinellid sponge 

 Ventriculites (a) has a very close resemblance, however, to the form of 

 calice under discussion. 



A further development leads to the flabellate type of Archseo- 

 cyathinse. This (as shown in figs. 7 and 8) is a large irregular expansion, 

 anchored by cylindrical bundles of rhizoid lamellae, and such structure 

 is unknown in the corals, so far as I am aware. On the other hand, 

 there are many sponges with a very similar form, such as the 

 tetractinelUd PhaJcellia (b). 



It is difficult to imagine that the two extreme types, turbinate 

 and flabellate, belong to two distinct families, as they have the same 

 fundamental structures of outer and inner perforate walls, connected 

 by perforate septa and tabulae. Considering the simple turbinate cup 

 alone (figs. 1 and 2) it would be easy to suppose it the skeleton of a 

 simple coral, uniting the outward form of the Rugosa with the septa 

 of the Perforate Hexacoralla. This has apparently been the view of 

 many earlier investigators. 



Where, however, did the polyp animal live ? One can imagine 

 an organism resembling the coral polyp, the gastral portion occupying 

 the central cup, and filamentous mesenteries penetrating into the 

 loculi, between the septa, through the (invariably) perforate inner wall. 

 But how can this hypothetical simple polyp be related to the sarcode 

 of the flabellate types ? The shape of the latter {ride C fig. 5) is highly 

 irregular — though the internal structure of walls and septa is constant 

 — and nothing resembling a cup or calice is present. One is led to 

 assume that in this latter (flabellate) type of Archaeocyathinae the 

 organism is of a compound nature ; and if this is true for the flabellate 

 form, it is most probably true for the infolded or turbinate type. 



A glance over the general scheme of Anthozoan evolution seems 

 to show a gradual advance from archaic solid dense types, such as 

 those obtaining in the rugosa, towards the delicate, and often highly- 

 elaborate trellised and trabeculate skeleton of the modern Madre- 

 poraria. It is somewhat like the advance from the solid strength of 

 a Roman aqueduct to the equally strong but more economical (and, 

 therefore, higher) type of the modern concrete-steel Monier equivalent. 

 Some of the etched Archaeocyathinae from Beltana reveal structures 

 more regulat-lv arranged and more strongly braced than any Paleozoic 

 or Mesozoic Anthozoan. Here, then, would be a striking example of 

 retrograde development if the modern madrepore were directly descended 

 from the Cambrian Archaeocyathinae. 



Reverting again to the simple turbinate type, let us consider the 

 " building construction " (to use a useful engineering term) of the 



(a) Lankester— Zoology, Pt. II., SpDnges, p. 16. (b) Ibid, p. 22. 



