BULLETIN 59, UNITED STATES NATIONAL MUSEUM. 195 



cance in fiO has the pains on this s<'i)tiiiii suppre.ssod. N'crv iicur thi- wsill iiiiinv or 

 most of the soptii split, and the forks are continued to the wall; either directive may 

 trifiircate. Over the point at which a septum splits is a dentation or granulate. 

 Tile portion of the septum l)etween this taller oranulati^ and the pulus mav hear 

 one or two minute granuliilioiis, these, however, may be absent, the portion between 

 it and the wall is excavated. A eircle of synai)ticuhi bind the outer ends of the large 

 granulates together; a .-ieeond circle of synaptit-ula often binds the pali together. 

 In center of the palar crown is the compressed, solitary lolunicllar tuben-le. There 

 is no well-developed cohnnella tangle. 



Loealifiei<. — Kaneoh(% Oahu, I specimen (type), \V. T. Brigham. collector; 

 AVaikiki. Oahu, Dr. J. E. I)uerclen. collector, tj specimens: Puktm. .Mojokai. Dr. 

 J. E. Duerden, collector. 



7;yy>,'.— Cat. No. 21t!27, T.S.N.M. 



JiciiKirkx. —T\n>i species is so entirely distinct from any of the other species of 

 /*w/<<'.v known from the Hawaiian Islands that critical notes are unnecessary. There 

 is a feature, however, that deserves especial consideration, namel3-, the bifurcation 

 of the septa near the wall. This condition suggests the genus Gonioporn instead of 

 Puriti's; but the bifurcation, or even trif urcation, of the directives is not Gonioporoid, 

 basing a judgment upon Bernard's remarks and diagrams in his "Goniopora."" Mr. 

 Bernard has kindly examined photograpiis of this coral for me and considers it a 

 true I'di'ites. 



The United States National Museum has received through the Carnegie Institu- 

 tion 7 additional specimens, collected by Dr. J. E. Dueiden. The calices of these 

 specimens show no noteworthy ditierence from those of the type, but there is 

 considerable variation in form. The corallum first forms an incrasting base, then 

 grows upward, becoming variously lobate. The lobes may })e more or less separate, 

 as in the type; the\' niay secondarily fuse and produce a glomerate upper surface; or 

 the corallum may be massive, increasing jn diameter as it rises above the base, have 

 ridges and depressions down its sides, and a glomerate upper surface. 



5. PORITES PUKOENSIS, new species. 

 Plate XCIV; Plate XCV, tigs. 1, 2. 

 Corallum forming thick, irregular, compressed or subterete, nodose columns, 

 on W'hich humps or stumpy protut)eranees mav occur. The columns rise from a 

 common base, and are more or less fused throughout their length, except the free 

 projecting ends, or in some instances they are fu.sed both above and below, leaving 

 intermediate open spaces. Two views, natuial size of the tyi)e s]iecimen, showing the 

 habitus arid size of the corallum are given on Plates XCIV, XCV. There are three 

 other specimens: The largest is of nearly the same size as the type, the columns dis- 

 tally diverge more, their ends are truncate, and some of them are more compressed. 

 One of the other specimens is young, incrusting a branch of a species of Porites 

 and sending up columns from 24 to -1:2 mm. in height, tapering to rounded or Iruni'ate 

 ends. The foui'th specimen is composed of several lobtvs. tapering to rounded ends, 

 and a twistetl, truncate plate, all I'ising from a conunon base. It shows no notable 

 difference fi-oni the tliird specimen. 



"The genus (!uniopor<i, Cat. Madrej). Corals, Brit. Miis. (Xat. Hist.), IV, UIO.S, p. I'l. 



