432 PROF. P. M. DUNCAN ON DEEP-SEA [May 16, 



long and granular at the sides ; and there are not quite three cycles. 

 The pali are broad and short, and look like continuations of the pri- 

 mary and secondary septa. The columella is small and concave. 



Breadth of calice •A- inch. Height of corallum T ^ inch. 



Locality. Dredged up from Persian Gulf on Nullipore. 



Paracyathus coronatus, sp. n. (Plate XXXVIII. figs. 1 1-1 3.) 



The base is broad, but not so wide as the calice ; and the whole is 

 short, the costse being very distinct, prominent, unequal, and extend- 

 ing to the base. The calice is elliptical in outline, and the larger 

 septa are exsert and rounded. The calice is shallow ; and a circular 

 ring of septa-like pali arises around the rather small columella. 

 The septa are in incomplete four cycles ; and there are about 20 pro- 

 longations into the inner circle. These pali are long and arched, the 

 whole presenting the appearance of an intercalicular gemmation. 



Height of corallum j% inch. Breadth of calice ^ inch. 



Locality. On a shell in the Persian Gulf. 



There is often much difficulty in deciding whether one of the 

 crowns of pali are really such, or only long spines attached to the 

 septa on their inner margin near the central space. 



The importance of deciding the true character of the structures is 

 great ; for whilst the septal spine may be of specific importance, the 

 presence of the pali as independent structures is generic ; for it in- 

 volves the presence of other tissues — such, for instance, as an extra 

 crown of soft tentacles. Every one who has seen many of the small 

 sessile corals usually called Astrangia, Phyllanyia, and Ulangia 

 amongst the Astrangiacese must have felt this difficulty. In some 

 the spinose nature of the false palus is evident ; but in other species 

 an arbitrary custom appears to have decided that such and such are 

 not septal structures but pali. 



In the Astrangiacese there must be evidences of endothecal struc- 

 ture in the form of dissepiments, although Milne-Edwards says 

 it is " peu abondante." Moreover the septa must have their free 

 margins more or less incised, and not plain, The origin of the coral - 

 lites from a basal expansion is part of the diagnosis ; but of course 

 this fails with regard to the parent before basal expansions, or stolons 

 have been cast forth. 



If a corallite simply increases by basal expansions, or stolons, it is 

 not necessarily one of the Astrangiacese ; for budding can take place 

 in species of other groups so low down and close to the base that 

 it appears to be, and may be, essentially basal. This is seen in a 

 specimen of Ovulina cubaensis from the Caribbean, when Ser}>ul(e 

 have kept the branches from rising as usual. Moreover in the 

 Sclerohelia from St. Helena there are some corallites close to the 

 base, and continuous with it, which are not distinguishable from it. 



Corallites springing thus from a base more or less closely, and not 

 having endotheca or serrate-edged septa, are not Astrangiacese ; so 

 that without the necessity of determining what are and what are 

 not pali, the difficulty in classification is somewhat removed. 



