115 



meagre description and more particularly the figures of which 

 lead me to doubt whether it belongs to the species at all. 



Caryophyllia clavus, according to Duncan and all other authors, 

 has five cycles of septa, whereas var. epilhecata in the figures, 

 supposed to be typical of it, is represented with a roimd calicle 

 and only four cycles. It is not a matter of size, since five of 

 Duncan's figures of C. clavus have their calicles of the same size 

 or smaller than the type, and yet have pali and septa of the 

 fifth cycle. Furthermore, I have examined a large number of 

 specimens of Duncan's var. smithi of 5 to 12 mm. in diameter 

 from English, Norwegian, and Mediterranean seas, and all with- 

 out exception possess a considerable number of septa of the 

 fifth cycle (quinary septa). 



The description of C. efithecata would be as follows : — 



" Corallum typically conical, often bent to one side and slight- 

 ly compressed, sometimes attached by its small end, sometimes 

 free owing to its peduncle having been dissolved away or 

 broken. Epitheca present, extending over the outside of the 

 corallum almost up to the edge of the calicle, and filling in the 

 intercostal spaces. Sub-equal costae present, showing through 

 the epitheca to the base of the corallite (if the corallum has not 

 been dissolved away) as low, broad, sub-equal ridges separated 

 by narrow valleys. 



" Septa forming four cycles, those of I. and II. sub-equal, 

 slightly broader and more exsert than those of III., which 

 again are broader and more exsert than those of IV. Pali only 

 before the tertiary septa, relatively thick, much-flattened rods, 

 of about 1*5 mm. in breadth. Columella a quarter to a fifth of 

 the diameter of the calicle in breadth, consisting of 2 to 9 

 twisted rods filling up the axial fossa." 



I append a table of the specimens, from which it will be seen 

 that 22 specimens have the typical number of septa, 4 specimens 

 have the fourth cycle incomplete by two or four septa, while 4 

 others show meristic variation, two having five, one seven, and 

 one eight systems. Eighteen of the specimens were, so far as 

 I could see, free, some showing scars of former adhesion and 12 

 attached, all except two, by quite narrow peduncles. Some 

 specimens show on their sides transverse ridges of growth, but 

 these are much obscured in many by the deposited epitheca and 

 in certain by both theca and epitheca having been dissolved 

 away, leaving the edges of the septa exposed with the inter- 

 septal spaces filled in, apparently by deposited corallum (stereo- 

 plasma). 



The size of the processes of the columella vary with the size 

 of the axial fossa, but more particularly with their own num- 

 ber ; in all cases they nearly fill in the fossa. The summit of 

 the columella likewise varies in depth, in some specimens being 



