THE SPONGES OF THE WEST-CENTRAL PACIFIC 155 



spaces as great as 7 to 14 mm deep. The endosome forms an exceedingly 

 thin incrustation attached intimately to the coral, following its irregularities. 

 It is about 80 /x to 600 /x thick and contains here and there very feebly de- 

 veloped fibers which do not seem to make any reticulation at all. 



The skeleton consists chiefly of smooth subtylostyles, about 6 /x by 

 500 fi. The heads are so very feebly developed that many of them might 

 properly be called styles. In some cases, they are arranged in tracts with 

 little or no spongin. The tracts or fibers are about 130 fx in diameter. Most 

 of these spicules are in the endosome ; very few, if any, in the conspicuous 

 but widely separated dermis. The microscleres consist of sigmas only slightly 

 curved, 33 /x in chord length ; toxas 40 fx in length ; abundant palmate aniso- 

 chelas, 40 [x long; and exceedingly rare palmate anisochelas, 15 fx long. The 

 microscleres are equally abundant in the dermis and the endosome. 



This species has a unique manner of growth. 



The name is derived from the Greek words for "roof" and "skin," 

 because the dermal structures make such a striking ceiling or roof over the 

 interior. 



GENUS OXYCARMIA, new 



This genus is established in the family Ophlitaspongiidae, to have as 

 genotype the species Oxycarmia confundata. It is characterized by mega- 

 scleres which include tylostyles, as typical of Carmia, and very numerous 

 oxeas, which is astonishing and probably shows relationship to the genus 

 Oxymycale, which is in the family Desmacidonidae. Microscleres, as in 

 Carmia, include anisochelas, toxas, and sigmas. The generic name is based 

 upon the resemblance to Carmia and includes reference to the occurrence of 

 oxeas. 



Oxycarmia confundata, new 



. Text Figure No. 101 



This species is here represented by the following : 

 U.S.N.M. No. 22874, My No. M. 170, here designated as type, collected 

 July 30, 1949, by diver in northwest Ponape between the reef and the 

 shore. The depth was 5 meters, and as substrate was growing upon a 

 live sponge, My No. M. 403, Neopctrosia pandora. 



The sponge is a thin crust, 1 mm thick, and covers about 10 square cm. 



The ectosome and endosome color in life was red, and the consistency 

 was mediocre. 



The surface is smooth and lipostomous. 



The ectosome is separated from the endosome, not only by being more 

 densely protoplasmic but by having many of the oxeas arranged perpendicular 

 to its layer. The protoplasmic dermis is only about 15 /x thick. In the endo- 



