liULI.ETlX 59, UNITED STATES NATIONAL MUSEUM. 137 



would not belong to the same f:iiiiily as SiderKKfivra. S/'dfraxfrea is a relative of 

 T/iamnaxteria," at least they possess rather similar septal structure. The principal 

 differences are that the septo-costje of the latter LTcnus are conHucMit and that eorallite 

 walls are absent. Sideriisfn'a belongs to the family Thamuasteriida'.'' The species 

 with which Pavona d^ierdeni groups belong to the Agariciidaj. 



The reference of the P. claviiff group of species to l\n'<niit now deserves further 

 consideration. For some time I was inclined to separate them from Panona and pro- 

 pose a new genus for them. The principal generic character would have consisted 

 in the continuous imperforate eorallite wall. The eorallite walls of Paronri are of 

 synapticulate origin. The same is true of PaKwia gigantea V'errill, which appears 

 to connect the species tj'pified bj' P. claviis with typical Pavona (type species, P. 

 crixtatd (Ellis and Solander)). I have not T)een able to study the development of the 

 wall in the massive Pavonw as carefully as is desirable, but in /'. gigantea 1 could 

 study it, and have already given its nature. The same is true of /'. duei'deni. The 

 wall is oi'iginally synapticuliite. the synapticula later forming a continuous imper- 

 forate wall. 1 am inclined to think that the same process will be found to occur in 

 the other species — this is, of course, reasoning from similarity in other characters. 



The P(iroii:v, in my o])inion. can be divided into three groups, based on mode of 

 growth, the frondose, the incrusting, and the massive species. Those that have just 

 been the object of special consideration belong to the massive species. 



Genus LEPTOSERIS Milne Edwards and Haime. 



1849. LeptoserU Milne Edwards and Haime, Coniptes rendus Acad. Sci., Paris, XXIX, p. 72. 

 1892. Folwueris Rehberg, Neue und wenig bekannte Korall., p. 26. 



Reliberg's FoUonerix is only a digitate Lcptoxerix. It is doubtful whether 

 Quelcli's Domoseris is generically different. Mi'. .). Stanley Gai'diiier says concern- 

 ing the relation of Domoseris to Li-ptoxerix : 



In fact, the oxanjination of the young forms [of Leplomris invrtmlans (iardiner] and that of tlie 

 original specimens [of Domoseris] causes one to suspect the generic separation of Domoseris Quelch 

 from tlu" trtinns under consideration [Leploseris']. •' 



LEPTOSERIS HAWAIIENSIS, new species. 

 I'latts XXXIN, XL. 



("orallum thin. The form is e.\tremely variable. The young eorallite attaches 

 itself, then it may grow into a thin funnel-siiaped corallum. or may extend itself on 

 all sides very nearly in a plane. The largest specimen has a greater diameter of 

 165 mm. The margin may be graduall}' curved, or may be lobate and crispate. 



Calices unifucial; a central calice can usually be recognized, the other calices 

 are irregularly scattered, frequently distant, occasionally more or less definitely 

 concentric when several are close together: but even then they are distinct, never 

 forming continuous valleys. They may be circular in cross .section or may be ellip- 

 tical, the longer axis transverse to the direction of the septo-costaj. The diameter of 



« Usually wrongly given as Thamnastriea. 



61 doubt the validity of a family Thatnnasteriidjf, but in tlie present uncertain state of our 

 knowledge it can be used in discussion. 



'Fungid Corals, Fauna and Geography of the Maldive and Laccadive .\rchipelagoes, p 948. 



