20 MADREPORAEIA. 



often melting down with the septa into a reticulum, and further they are septal structures, and 

 in matter of mere bulk frequently insignificant as compared with the parts contributed by 

 the more radial portions of the septa ; hence, as stated above, we call the wall, the whole of 

 the skeleton which forms the rim of the theca. 



The synapticulas may form a single ring as in the case just cited. This may be uniform 

 or it may show its composition out of so many separate synapticulas by running in a zigzag 

 round the calicle. If the synapticulffi do not come quite to the edge of the wall, tlie latter 

 is denticulate, or, if the wall is thick, it is striated by septal ridges, these ridges being of 

 course the homologues of the denticulations of the thin walls. 



Lastly, the wall may lose all regularity in the arrangement of its radial and concentric 

 elements, and may melt down to form a reticulum, and this reticulum may have many 

 characters. It may be delicately filamentous, PL II. fig. 9 ; or composed largely of flakes. 

 When these flakes lie horizontally we get such an extreme form as that shown, PI. V. fig. 5 ; 

 when they are arranged vertically we get a peculiar type, examples of which are shown in 

 PI. VII. fig. 1, PL VIII. fig. 1 ; in which the walls are composed of twisted vertical lamellas. 

 Many grades occur between these three extremes, the purely filamentous and the horizontally 

 and vertically lamellate. 



Other modifications of calicles are bound up with variations in the septa. But it may 

 here be noted that when the walls melt down into a reticulum the septa are usually involved 

 throughout their whole lengths, and we may have calicles in which all radial symmetry is 

 obscured, the whole skeleton being little more than an irregular sponge-work * (e.g. PL VI. 

 fig. 1). One modification on the primitive type occurs abnormally on a specimen from Celebes. 

 It is figured PL IX. fig. 3. So far it is not known on any representative of the genus 

 as its normal calicle. 



The Septa.— The primitive calicle had 24 septa, which still persist in the great majority 

 of the modern representatives of the genus. Even in those forms in which the calicles are 

 most specialised and the number of septa reduced and .their symmetry obscured, the typical 

 number will usually reappear in the " lateral " calicles, or in those on " free edges," which, as 

 above noted, tend to revert to the primitive type. 



In certain fossU forms, however, in which we have no lateral calicles preserved for us to 

 turn to, it is often extremely difficult to say whether there are 12 or 24 septa. We may have 

 for instance 12 distinct plates radiating outwards from the centre and round these apparently a 

 very thick reticular wall. It is difficult to decide whether this reticulum is all mural or due 

 to the secondary fusion of the septa, the tertiaries coming in to complicate matters. In these 

 cases there is only one certain method of deciding, and that is not always available. When a 

 median continuous wall- or synapticular thread runs through this reticulum round the calicle 



* Cf. the remarks made p. 146 on the Paris Basin forms. It is there suggested that the 

 melting down of the skeletal elements, septa, and synapticula; may be a protection against deposits 

 of fine mud which might otherwise settle down into the interseptal pockets of the skin in retracted 

 polyps and be difficult to dislodge. 



