of some Corals from Fiji. 217 



name 8. raristella, var. wilsoni, after the gentleman, who besides 

 sending the specimens is largely responsible for the information 

 about Levuka harbour. The variety differs from the type in 

 having the ring round the calicle less distinct, but crowned with 

 the six large spines of the primary septa, generally six smaller 

 intermediate spines, and often six or twelve of a third order, no 

 trace of any corresponding septa however beyond the first order 

 being ever visible. The calicles tend to be disposed in series, 

 which do not lie necessarily in any determinable direction in 

 respect to the branches ; on the tips of the latter the series are 

 ordinarily especially marked, crossing them in parallel rows. The 

 coenenchyma varies considerably, but is fairly well developed, 

 and covered with low pointed spines ; the calicles of a series at 

 the tip of a twig generally have a common theca between, but 

 below all tend to be separated by a breadth of about their own 

 diameter (1 mm.) from one another. Often a row of spines divides 

 the coenenchyma markedly between two separate calicles, or two 

 series. The septa have smooth edges, and fuse below with the 

 columella. The latter is always distinct above, projecting freely 

 generally to the level of the top of the " bourrelet," or rim round 

 the calicle. 



Two of the other specimens evidently belong to the same 

 species, which I referred to Pocillopora suffruticosa Verrill, in 

 describing my collections from the Pacific Ocean 1 . The identi- 

 fication down to the species is however uncertain at present, but 

 it serves to indicate that the specimens belong to the finely 

 branching division of the genus, noted also for its dense corallum. 

 The last specimen may be provisionally referred to Pocillopora 

 plicata (Dana) ; it is too small to determine properly its mode of 

 growth, but it approaches close to var. aspera (Verrill) 2 . 



In any discussion of the rate of growth of corals the volume 

 of the specimens must be considered rather than the weight, since 

 according to all my observations the density of all forms of a 

 single species varies inversely with the rate of growth ; the slower 

 the latter may be, the denser and heavier is the corallum, volume 

 for volume. In the appended table I give the particulars of the 

 rate of growth of the five specimens. In the third column I 

 record the weights of the colonies, when they first arrived in this 

 country, about three months after being obtained, to show by 

 comparison with the fourth column the additional loss by thorough 

 drying. In the last column I give the thicknesses of the coralla, 



1 P. Z. S., 1897, p. 943. 



2 P. Z. S., 1897, pp. 947—8. In looking up this reference I find that I fell into 

 error in calling P. aspera the type. It is indeed the central form round which the 

 varieties are grouped, but according to the recognised rules of nomenclature the 

 type should be P. plicata and the three varieties var. danae, var. ligulata, and var. 

 aspera. 



