200 CEYLON PEARL OYSTER REPORT. 



The calicular fossa is rather narrow, and the septal trabecular forming the columella 

 si tort and bluntly spiniform. This specimen might be referred either to F. rubrum 

 or F. stokesi, M. Edw. and H. ; but it cannot be either F. pavoninum or dLstiactum, 

 the colour of the septa notwithstanding. These three specimens, therefore, are 

 varieties of F. rubrum, QuOY and Gaimard, which is synonymous with F. variabile, 

 Semper, but is, in my opinion, distinct from F. crassum, M. Edw. and H. (= F. irrc- 

 gulare, Semper). 



I differ, with some hesitation, from Mr. Gardiner, but I feel bound to point out 

 that his reasons for uniting nearly all the described species of Flabellum under 

 F. rubrum are not satisfactory. He relies very much on certain numbers and 

 measurements, and he appears to think that if one form is connected with another by 

 some individuals, the two must be reckoned as constituting one species. Now, in the 

 first place, much depends on the numbers or measurements that are chosen for 

 comparison. Two of his characters, viz., the number of septa fusing by trabecule 

 and the total number of septa, are characters depending on the age of the coral, and 

 are therefore of little value. His third character, the relation of the length to the 

 breadth of the calicle, is tolerably constant at all ages after maturity, and is therefore 

 better. But whatever characters are chosen, it is not sufficient to set out the results 

 of the measurements in a simple table and to say that, since the average measure- 

 ments of a number of unequal groups of individuals can be arranged in a continuous 

 series, all the specimens measured must belong to one species. To deal with the 

 statistics of a number of individual forms, proper statistical methods must be 

 employed, or systematic zoology will be thrown into confusion. The extreme 

 measurements the characters that vary most widely from the mean of two closely 

 allied species may be expected to overlap, but the fact of their overlapping does not 

 break down the distinction between two species whose mean is different. We have 

 at present no data for determining the range of variation in the species of Flabellum 

 by statistical methods, and until such data are available I prefer to place reliance as 

 much upon such characters as the shape of the columellar trabecular, the shaj>e and 

 thickness of the septa and the kind and arrangement of the granules on their 

 surfaces, the visibility or otherwise of the costas, &c, as upon measurements which 

 have been shown to vary very widely according to the age and condition (whether 

 recently liberated or not) of the individual. 



Placotrochus, M. Edwards and Haime (1848). 

 The reasons for placing this genus in the family Flabellidse have been given above. 



Placotrochus laevis, M. Edw. and H. Plate I., fig. 5. 



Three specimens from Periya Paar are referable to this species. Two were 

 preserved with the polyps partly extended, one is dead and corroded. The three 



