14R MADREPORARIA. 



arc low, thick, round -topped, but not conspicuously reticular, composed rather of the thick, 

 solid, peripheral ends of the septa. The septa appear to be uniform all round the calicle, and 

 to slope sUfjlitly from the walls, where they are wedge-sliaped with conspicuous granules along 

 their edges, and so lose themselves in the open, flat base of the fossa which is occupied by a 

 mass of granules without apparent arrangement ; close inspection, however, shows them to be 

 especially large at the points of fusion of the septa, which can then be seen to be arranged in 

 the typical manner. 



Shallow calicles with a tendency to have the edges of the septa brol<en up into granules 

 are characters belonging to very thin explanate forms, as already described in Vol. IV. Three 

 other examples are already known of this, viz. G. Maldives 1 (see Vol. IV. PI. VII. fig. 1), and 

 G. North- »''cst Australia i3 (PI. IV. fig. 1), and G. China Sea 'J (PI. V. fig. 4). 



The specimen is not only interesting because it supplies us with another typical instance 

 of this method of growtli, with its characteristic calicles, but because it is encrusting a dead 

 fragment of an entirely different kind of Goniojmra, see below, G. Ellice Islands 3, spec. /'. This 

 latter point claims attention in a genus which does not seem to be at all common. It shows 

 tliat in one and the same locality we may have forms quite distinct from one another living 

 side by side. 



a. Zool. Dept. 1903. 4. 3. 2. 



155. Goniopora Ellice Islands (4)2. {0. Elliciana secunda.) (PL VIII. fig. 1.) 

 [Funafuti, coll. W. J. Sollas ; British Museum.] 



Description. — The corallum apparently started as a low, convex mass, encrusting the 

 worn and loose fragment of a branching coral. TMs latter apparently rolled over, and the 

 Ooniopora, with great distortion of the calicles, has had to grow up round it, with the result 

 that it appears as if it had been forcibly wrapped round the fragment to which it is attached. 



The calicles are sliallow and open, like those of encrusting fonns, ranging from 2 to 3 • 5 mm. 

 across. The walls are low, thick and well marked, composed of a coarse, irregular, rather flaky 

 reticulum, the flakes running out into free ends. The septa, wliich show clear traces of the 

 typical formula, are thin, jagged and very short, Ijecause they soon join the large columellar 

 tangle which is of the same flaky reticulum as the walls. 



At the point of most rapid growth, the calicles are a little deeper, the wall reticulum a 

 little more open and filamentous, the septa, of which some eight are conspicuous as thin plates, 

 are reduced, l)y the large size and irregularity of the perforations, to filamentous lacework. 

 The columellar tfingle is smaller and more openly reticular. 



Here and tliere the wall reticulum and the columellar tiingle run together and obscure 

 some of the se]ilii. 



This coral is clearly an abnormal specimen, and it is impo.ssilde to say what its calicles 

 Would have been like had it found a stable body to rest upon. 



The tendency of the skeleton everywhere to ])ecome reticular is obviously pathological, 

 for it goes to the greatest extreme in that part whicli suflered most by the rolling over of its 



