34 



PROCEEDINGS OF THE NATIONAL MUSEUM 



VOL. 81 



P. laminans. This shape is found, however, mainly in the largest 

 fragments, some 10 cm in diameter, which seem to have been part 

 of a large platelike growth about 20 cm high; its oscular surface 

 covers most of the concave side. On the small fragments such oscular 

 fields as the following are observed : Oval, 1 by 2 cm and depressed 

 1 cm; hourglass shaped, each oval 1 by 2.5 cm and depressed 1 cm; 

 oval, 1 by 3.5 cm and depressed 1 cm; triangular, 1 by 1.5 cm and 

 depressed only about 3 mm. It must be stressed that each oscular 

 area was surrounded by a dense coronal. palisade of very long spic- 

 ules, usually very close to 17 mm in length except in the smaller 

 specimens. The function of these seems to be separation of exhalent 

 current from inhalent. 



On May 1, 1929, Mr. Kicketts brought up a macerated hexac- 

 tinellid dictyonine skeleton from very much the same depth and 

 locality as of the type. On it were about a dozen sponges that I iden- 



FiGORE 13. — Poecillaatra rickettsi de Laubenfels : A to F, X14; others X300 



tify as conspecific here. These are rough cylinders, the largest about 

 1.5 cm high and 2 cm in diameter, the smallest 1 cm high and 3 mm in 

 diameter. The spiculation is the same, except that I do not find the 

 toxas in them. The oscular areas with covering fenestrated mem- 

 brane, the convex pore surfaces, and the dense coronal palisades are 

 all the same. The inference is that this species gTows to a height of 

 about 1 cm before expanding laterally very much, but that after- 

 wards its ijrowth is almost exclusively horizontal. 



Do all Poecillastras have such a change of form during their life 

 history? It is to be noted that many of the specimens might fairly 

 be classed as jSphinctrella, except that they lack the annulation so 

 characteristic of the spicules of those certainly SphinctreUa. 



Were the toxas proper, they would be a most striking feature. As 

 it is, the enormous oscular crowns distinguish this species from all 

 others of the genus. P. laminans Sollas (1886, p. 186) from the 

 East Indies is probably the closest, as it had a low (4.5 mm) fringe 

 around its oscular areas. 



