BULLKTIN 59, UNITED STATES NATIONAL MUSEUM. 81 



the name, and also know that it was against the rules of nomonclature to inaiie a 

 species itisertcd into a genus suhsecjuent to its original characterization the type- 

 species, he decided to follow the usage established by Lamarck. Dana, Milne Kdwards 

 and Hainie, and Duncan. In my Some; Fossil Corals from Curayao, Arube and 

 Bonaire", I abandoned this use of the name and proposed substituting hopora, 

 Studer, 1878, but did not attempt to fix the type of the Linna>an Mdilrcjuirii. Professor 

 Verrill* pointed out that ^icrn/x/rn Oken, 1815, must be used for the Lamarckian 

 Madrepora^ and that following Oken's restriction of the Linnsvan Madrepura, either 

 M. jtroliferd or J/, octdata. both of Linnanis, could become the genotype. For 

 several reasons preference is given to M. oculata. Therefore Jladrepora oculata 

 Linnsus becomes the type of Madrepora Linnteus. 



MADREPORA KAUAIENSIS, new species. 

 Plate VIll, lifrs. 1, 2, 'la. 



Corallum witli delicate branches. Asexual reproduction bj' budding from below 

 the edge of the calice; rather often on each side of a calice, leaving the mother calice 

 more or less inmiersed in the angle between the daughter calices. Such a sunken 

 calice always occurs in the axis between two branches. When no bifurcation takes 

 place gemmation is alternate, forming a row of calices on each side of the branch. 

 Terminal calices are prominent, as nuich as 3.5 nmi. in height, the corallite gradualh' 

 enlarging toward the aperture. The lateral calices on young Ijranches are moder- 

 ately prominent; on older branches the coenenchyma ma}' extend upward and leave 

 onl}- a small portion of the corallites free. 



Around the upper outer margins of j'oung corallites there are distinct, often 

 acute, costse. Farther down on the wall the}' are low, but can be traced. They are 

 minutely and densely granulated. On older portions of the corallum these costse can 

 not be traced, but Hexuous, often coarse, c(enenchymal striations are present. The 

 whole coenenchymal surface is minutely granulated. 



Septa in older calices in three complete cycles, the third cycle well developed; 

 in yoiuig calices the last cycle is rudimentary, but traces of it can be seen; septal 

 margins entire. The young calices are verj- deep, but the older ones are shallow, 

 the hotfornK I if the c<d'ice» Ijecomhuj xolidly filled with utrreoplasin. 



Columella very poorlv developed in 3'oung calices, consisting of a few trabec- 

 ule, which appear to be derived from the inner ends of the septa. The upper 

 surface is usualh' papillate. In older calices it is much bettei- developed and may be 

 considerably con)i)acted In- calcareous deposit. A few outstanding papilla? often 

 simulate pali. 



Locality. — Vicinity of Kauai Island, Stjition WM\ depth, 21>-i-352 fathoms; 

 bottom, fine coral sand; temperature, 44. 'i"-" F. 



TyjOf.— Cat. No. '20780, U.S.N.M. 



Rewarkx. — Duncan in his first paper on the Porcupine Expedition Madreporaria'" 



«SammI. Geolog. Reichs-.Mus., Leiden. 2d Ser., II, Hft. I, p. 68. 

 ''Trans. Conn. Acad. Sci., XI, 1902, p. 110. 

 < Proc. Koy. 80c. London. XVIII, 1870, p. 295. 



H2301— 07 6 



