18 MADREPORARIA. 



living tissue. This is probably more especially true of the porous forms, for a iinely 

 reticular skeleton is obviously better adapted to follow the modification of living tissue 

 than one constructed of stiff plates. These remarks are, however, thrown out tentatively, for 

 little as we know about the method of the secretion of the skeleton by the ectoderm, we know 

 absolutely nothing of the physiological stimulus which causes it here to secrete and there not 

 to secrete, a differential action which leads to all the varying patterns of the skeleton. 



The fundamental character of the Skeleton of Goniopora. — The skeleton is porous because 

 constructed on the same general principle as that of the Madreporidte. The primitive 

 epithecal cup or prototheca of the Madreporaria has been flattened out, and become vestigial, 

 and the present thecpe are entirely septal and internal.* 



This internal skeleton is usually described as " trabecular," that is, its principal elements, 

 the septa, are supposed to be built up of trabecule or skeletal rods arranged as a lattice-work. 

 This, however, is not a very accurate description. It is true that the septa are usually very 

 perforate (see PL III. fig. 2) but nothing is gained by describing a perforate lamella as 

 being composed of the strips of tissue running between the perforations ; for it is the septa 

 which are the morphological units and not these intervening strips. The term trabecula is, 

 however, too useful to be altogether discarded, and it is here used in the same sense as in 

 Vol. III., viz. in descriptions of texture, sometimes to denote the thick and prominent skeletal 

 strands which run in the direction of growth, the tips of which, appearing at the surface, 

 e.g. on the tops of walls and along the upper edges of septa, form granules and spikes. 



The Vestigial Fjntheca. — In accordance with the origin here assigned for this family, viz. 

 as due to the fixing of a perforate i.e. purely septal skeleton at an immature stage when the 

 theca is low and shallow, we should expect to find more conspicuous traces of the epitheca 

 than in the Madreporidpe proper. For a vestigial structure would naturally be a more con- 

 spicuous element in a rudimentary stage of skeletal development such as we have in Goniopora 

 than in the case of a skeleton which towers up, and whose buds appear high up above all 

 contact with it. Explanate Madreporidie have theii- free edges usually shooting far beyond the 

 supporting epitheca. But in Goniopora and in Porites, the supporting epitheca takes the lead 

 in the formation of the edge and is frequently seen projecting, e.g. PI. III. fig. 6. 



The growth-forms of Goniopora are, for reasons which will be explained below, very rich 

 in free edges ; consequently bands of wriukled epitheca are a conspicuous feature in most 

 colonies. See further below on the tabulse, which are also related structures. 



The Theca. — Above the flattened epitheca rises the new theca or wall built up of the 

 septa united by synapticulse, and hence necessarily porous. In the majority of cases it is not 

 easy to define which part of this porous mass constitutes the wall, because there is generally 

 more than one ring of synapticulse, and even when only one ring is developed, the septa 

 usually carry incipient synapticulte whose further growth would immediately thicken the theca 

 and change it from what is here called a " simple " into a " compound " wall. 



The statement of Dr. Klunzinger, repeated by Martin Duncan (see Historical Eeview, supra, 



* Cf. Journ. Linn. Soc, xxvii. p. 487. 



