I'JS KKCKNT MADKKI'ORARIA OF THE TTAWAITAN ISLANDS AND LAYSAN. 



of gt'iierii' iiii.l family valiio. The fiiiiiilics of the Fiimj'idn have been based on iimral 

 uiid septal structmc. In my paper jii.st quoti>d. I remarked: 



The larger divisions an? based upon septal structure; that is, whether the septa are solid or 

 perforate; if perforate, whether they are more pronouncedly laminar or tral>ecular in composition, 

 and I have also utilized in deliniuf; the families the character of the wall, whether normally perforate, 

 even if only slijrhtly, or wlu'tlu-r normally solid. 



Sutiii'ieiit evidence lias, 1 helievc, In-eii ])resented to show that species ranging 

 from Fidui'ki (■le(jiin.'< Verrili, with a solid wall and solid septa, to sjiecies like 

 F. distorfa, whieh has a solid wall, l)iit extremely cril)iiform septa, and F. fimgites, 

 which has a very perforate w;dl and more or less perforate septa, form a continuous 

 series within oeneric limits. This series shows that the solid or perforate wall for 

 these corals is not of oeii(>iic, nmch less of family, value. F. eJegans Verrili is a 

 typical member of the Lophoseridie (Aoariciidie), if the pecidiar mode of asexual 

 reproduction from the tro])hozooid is left out of consideration. The size of the 

 septal dentations posses.ses ni> value as a generic character. The septal structure 

 varies from imperforate to extremely cribriform, covering four other families of 

 the Fiui(/!(la. namely, the Agariciida-, with solid septa, tiie Thanmasteriidic and 

 Leptophylliida', with septa that are solid or perforate, but which are always more 

 pronouncedly lamellar ihan trabecular, and the Anabraciida^, whose septa are 

 comi)osed of a tiiiliecular lattici'work. 



This study throws doubt on all the attempts to subdivide the Fungida into 

 families. For the family l''ungiidie, the only difi'erential character left is the mode 

 of reproduction. Should its sui)pos(>d value be proven erroneous, the consideration 

 of th(> Fungiida' as a family s(!parate from the other Fungida nuist be abandoned. 



Th(> data ])resented in the foregoing discussion are suggestive and not conclusive. 

 They emphasize, however, the need for the caref id comparative study of large series 

 of species to deternnne what characters are of actual value in establishing the higher 

 systematic divisions. 



Application is here made of the knowledge of septal structure derived from this 

 study of Funyiit by placing in Leptoseris & species with perforate septa next to one 

 whose sejjta are solid, and as a result oidy two fannlies are recognized among the 

 Hawaiian Fungida. 



FUNGIA PATELLA (Ellis and Solander). 



I'late XXVII, fijrs. 2, 2ii, 3, 3-(. Plate XXVIIl, titrs. L', 2a. 



1786. Madreimrii patella Ei.Lis and Sol.axder, Nat. Hist. Zooph., j). 148, jil. x.wiii, tigs. 1— I. 

 liH12. FuKijUi jHilella Doderi.ein, Senckenb. naturfor. Gesellseh. Alihandl., XXVII, p. 65, pis. i, 

 n (all liffs. ); pi. v, iigs. 1, 2 (Synonymy.) 



The Alhatrom obtjiined a large number of specimens of this species, and these are 

 used as a basis for the following description and table: 



('oralhun ncaily circular, slightly elliptical, somewhat arched above, sid)plane or 

 excavated in central portion; central scar indistinct or not visible. 



