188 CEYLON PEARL OYSTER REPORT. 



fossil species has in several cases been overlooked. Hence one cannot deal with any 

 collection of Turbinolida3 and Eupsammiidse, however small, without entering upon a 

 criticism of these families, and, unfortunately, my claims to undertake such a 

 criticism are limited. A short study of the literature of the subject is sufficient to 

 convince one that no sufficient degree of exactitude can be attained without a study 

 of the original specimens on which the genera and species were founded. But these 

 are largely contained in foreign museums, which I have been unable to visit, or were 

 included in private collections which have been dispersed and are no longer traceable. 

 The British Museum is rich in the possession of the " Challenger" collections and in 

 duplicates of species collected by Pourtales, but it is singularly poor in Eupsammiidae, 

 and a careful search among the named specimens of this family failed to bring to light 

 anything that was of assistance in disentangling the difficulties attendant on its 

 classification. I have, therefore, had to rely chiefly on published descriptions and 

 illustrations, and these are, in many cases, too vague and inaccurate to be of material 

 assistance. 



I wished, when I undertook this piece of work at Professor Herdman's recpiest, to 

 make a thorough examination of the anatomy of the soft parts as well as of the 

 coralla of the specimens he entrusted to me, but this I found was impossible, as 

 several of the more important species were represented only by a single specimen and 

 it was necessary to dissolve away the soft parts in order to identify the coralla. 

 T have, however, made a tolerably thorough investigation of the anatomy of three 

 forms, Heterocyathus cpquicostatus, Heteropsammia rnichelinii, and Dendrophyllia 

 gracilis. As a matter of convenience this paper will be divided into two sections, the 

 first dealing with the systematic description of the coralla contained in Professor 

 Herdman's collection, the second with the details of the anatomy of the three above- 

 mentioned species. 



PART I.-SYSTEMATIC. 



Family : TUEBINOLID^E (pars.), M. Edwards and Haime* 



M. Edwards and Haime divided the Turbinolidae into the two sub-families 

 Caryophyllince. with one or more crowns of pali, and Turbinolinae, without pali. The 

 Caryophyllinse were further subdivided into Caryophylliacese, with a single circle of 

 pali, and Trochocyathacese, with more than one circle of pali. The Turbinolinse were 

 subdivided into the Turbinoliacea?, destitute of epitheca, and Flabellacefe, in which 

 " the wall is completely covered by a pellicular epitheca." Martin Duncan, in his 

 revision of the families and genera of the Madreporaria (13), abolished the sub-families 



* 'Ann. des Sci. Nat.,' 3 e ser., t. ix., 1848, and 'Hist. Nat. des Corall.,' t. ii., p. 7, 1857; 

 P. M. Duncan, "Revision of the Madreporaria," 'Journ. Linn. Soc.,' vol. xviii., "Zoology," 1885. 



