504 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 14 



and the proximal border a little more arcuate), and the row of pores 

 around the distal side of the aperture are all much like those of 

 T. porosa. There is, however, a striking difference in appearance of the 

 species due to the presence of 4 (2 to 6) strong, pointed, erect spinous 

 processes around the aperture; these sometimes bear minute avicularia 

 at their tips (as in T. turrita Smitt) but usually they are either strong 

 tubercles or end in sharp points. The frontal is often somewhat roughened 

 and the pores obscured. Rarely a minute suboral avicularium is present 

 and very small rounded ones occasionally occur on the frontal. 



The ovicell is smaller than that of porosa, about 0.35 mm wide, the 

 primary layer with small pores, but this soon becomes covered with the 

 rough secondary layer, leaving temporarily a small central porous area, 

 but this eventually also becomes closed and a pointed umbo may develop 

 on the top. 



There is considerable resemblance to T. turrita but hexagonalis is 

 much smaller, more vitreous in appearance, the frontal is more porous 

 and there are no larger spatulate avicularia. 



Described by Canu and Bassler from the Galapagos Islands, Alba- 

 tross D.28 15. 



Hancock Stations: 438, Chatham Island; 450 and 155-34, Albe- 

 marle Island ; 452, Charles Island ; and 810-38, Barrington Island, Gala- 

 pagos. Also at 267, Angel de la Guardia Island, Gulf of California, 

 and 491-36, Rosario Bay, west coast of Lower California. In the col- 

 lections there are also specimens collected at Banderas Bay, Mexico 

 (George Willett) ; off Acapulco, Mexico (F. E. Lewis) ; and West 

 Mexico (H. R. Hill), through the courtesy of the Los Angeles Museum. 



The known geographic range is from the Galapagos Islands to about 

 30°N Lat., and the bathymetric range from 5 to 75 fms. 



Genus COSTAZIA Neviani, 1895 



Genotype, Cellcpora costazii Audouin, 1826. Until rather recently 

 the species of this genus were allocated to the old Linnaean genus 

 Cellepora. Neviani's description was apparently overlooked until Canu 

 and Bassler reestablished it in 1920. Waters in 1889 to 1913 confused 

 It with Lagenipora Hincks, and Levinsen in 1909 erected a new genus, 

 Siniopelta, with C. costazi as the type. The group of species is now 

 well enough understood to indicate its distinct separation from any of 

 the other celleporid genera; also the nature of the front and of the 

 ovicell distinguish it at once from Lagenipora, which evidently does not 

 belong with the Celleporidae. 



