GONIOPORA. 19 



p. 6), that the walls are united directly, and not by coenenchyma, has little real meaning. 

 Ccenenchyma is the porous tissue between the adjacent calicles of porous corals and these 

 united walls themselves form the coenenchyma. It is quite wrong to regard the fusing walls 

 of a Madrepore as ccenenchyma merely because the rims of the calicles tower above the level of 

 fusion, while looking on the fusing porous walls of Gonuqwra or Porites as only fused walls 

 and not coenenchyma. The coenenchyma in Goniopora is reduced to a minimum when the 

 wall is a mere perforate membrane, and it reaches a maximum when the fossae are most widely 

 separated by porous tissue. 



We may call the wall that portion of the theca which rises the highest, for there the synap- 

 ticulffi are always found uniting the septa in one or more rings forming respectively simple 

 or compound walls. We may divide the thecie into two types, regular and irregular. 



The regular theca is that in which the septa are symmetrically radial, and the synapticulae 

 more or less regularly concentric. This regular type may be regarded as the primitive type of 

 the parent calicle of the genus. 



Under the irregular types come all the other forms of calicles due to irregularities in the 

 radial arrangement of the septa, or in the concentric arrangement of the synapticula?. which 

 are found witMn the genus and which can all be deduced directly or indirectly from the 

 regular or primitive type. 



Tlic Primitive Calicle. — In addition to the regular radial arrangement of the septa, the 

 other chief characteristics of this parent form were as follows. The theca was low and tliick, 

 the fossa shallow, the septa fused according to a formula presently to be described, and the 

 columellar tangle was conspicuous in the floor of the fossa, and from it a rosette of six paliform 

 knobs arose. This primitive type of calicle is that which was supposed, on account of this rosette, 

 to distinguish the genus Rhodarcea, but as it plays an important part in the detailed description 

 of nearly every representative of the genus it could not possibly form a generic distinction. 

 It reappears almost invariably on what are called the " free edges " of colonies, no matter how 

 specialised the calicle structure typical of those colonies may be, so that all massive Gonioporas 

 are Rhodarseas at the sides ! The significance of this tendency of laterally placed calicles to 

 revert to the primitive type will be discussed in the paragraph on the Growth- forms. 



Modifications on this Primitive Calicle. — The theca being built up of septa united by 

 synapticula", varies with any variation in either of these two factors. 



If the top edges of the septa are all on a level, there is then no raised wall, and the 

 caUcles are said to be " obsolete " or flush with the surface. But in the great majority of cases 

 the wall is raised, and this raising reaches an extreme when one thick trabecula rises up 

 from the peripheral end of each septum, and is iinited to its fellows by broad band-like 

 synapticuhe, so that the wall is tall, membranous, and perforated.* (Of. e.g. PI. VIII. fig. 4.) 



The synapticulse joining the wall portions of the septa should perhaps be called the wall 

 proper, being the concentric elements. But they are not always symmetrically concentric, 



* The tall thin wall was the charactei'istic of the limited genus Goniopora M.-E. & H., I85I 

 (' Les Poritides,' p. 38). 



D 2 



