MADEBPOEAEIA OP THE DEEP SEA. 325 



on the Corals of the First 'Porcupine'" Expedition, respecting the very close alliance 

 of AmpMhelia and Lo])hohelia consistent with facts I have observed since. I find that 

 there are never any dissepimental structm-es developed in Aniphihelia, whilst they exist 

 and even take up the tabulate form in many specimens of Lophohelia. 



The absence of the columella, in all the specimens of AmpMhelia 1 have seen, is very 

 rare indeed ; and as the structure is never observed to be wanting in many consecutive 

 corallites on a stem (for even at the worst the large septa join centrally), the want of 

 the organ may be an arrest of development. 



Every now and then the third cycle of septa is so very rudimentary in AmpMhelia 

 oculata, Linn., that doubts may be expressed about its existence; yet in the same 

 corallum it may be visible in remote calices. 



The existence in AmpMhelia of well- developed spongy and trabecular columellae, ot 

 very small columellae of the same consistence, and of those formed by a swollen state 

 of the inner and adherent margins of the larger septa is beyond doubt. 



The ornamentation of the costse and their general development afford specific 

 distinctions. 



By grouping the specimens obtained from the first and second expeditions of the 

 ' Porcupine,' the following series may be distinguished : — 



1. Columella moderately developed. 



Septa in two cycles, first and second equal, the third cycle either very rudimentary 



or wanting. 

 Surface of wall plain. Costte very fine, and serpentine near the newest calices. 



2. Columella very small. 



Septa in three cycles, first and second unequal. Septa dentate. 

 Surface plain. 



3. as no. 2. Surface minutely and sharply granular. 



4. as nos. 2 & 3. Costae fine and visible near the calices. 



5. Columella moderately developed. 

 Septa in three cycles. 



Surface marked by continuous costal striae. 



6. Columella very small. 



Septa in three cycles, tertiary dentate. 



Surface marked by continuous and noncontinuous costal striae, and covered with 

 granules. 



7. as no 6. 



Surface strongly striated. Septa not dentate. 



8. Columella small. 



Septa in three cycles : the primaries are often exsert. 

 Surface of wall plain. Costal striae near new calices. 



' Proo. Koyal Society, 1870, P. M. Duncan. 



