ART. 12 HEXACTINELLID SPONGES — OKADA 55 



and broad at the base, measuring 8jx, attenuating gradually toward 

 the pointed ends. The surface is faintly rough. 



The more robust onychasters, measuring about 50/x to TO/i in diame- 

 ter, are also found together with the spicules on the upper side, but 

 more numerous in the subgastral or in the subdermal regions. All 

 the rays are somewhat slender and inconspicuously roughened near 

 the ends. The terminations of the branch rays bear a verticil of 

 fine claws, usually four in number. These are perpendicular to the 

 branch ray and slightly recurved. 



Small onychasters are present in the parenchyme layer, intermixed 

 sparsely with larger onychasters. They are more abundant in the 

 subgastral regions. They measure 30/x to 40/* in diameter and have 

 two or four widely diverged terminals, the surface of which is quite 

 smooth. 



In the compact, thickened regions of the sponge wall near the 

 sieve plate, the dermal hexactins occur more densely than in other 

 parts of the entire sponge body. The proximal ray of the hexactins 

 is much longer, attaining a length of 680|Lt to 750/*, while the distal 

 poplarlike ray is somewhat slenderer and shorter, measuring I6O/1 

 to 200/A in length. The tubercles of the proximal ray are very pro- 

 nounced toward the end of the ray. In general, the distal poplar- 

 like ray of the spicule is much broader and stouter, being much longer 

 than that of the ordinal hexactins distributed in other regions of 

 the sponge body. In the parenchyme of this region, the microscleres 

 are rarely found ; especially lacking are the hexactinic and hemihex- 

 actinic forms and the onychaster. 



The diactinic gastralia are thickly accumulated and irregularly 

 distributed, being many times as thick as the ordinary ones of the 

 subgastral regions, and measuring 0.8 mm to 1.7 mm thick. (It is 

 somewhat interesting to note that a dictyonal framework is not 

 found in the parenchyme of this region.) 



Uncinates, when they occur, are also found in fewer numbers. 

 They penetrate vertically to the surface, nearly approaching the 

 gastral surface. 



The peculiar oxyhexasters occurring on the gastral membrane are 

 practically absent, and when they do occur they are sparsely scattered. 



The so-called latticelike plate of this species differs somewhat from 

 that occurring in Aphrocallistes heatnx Gray in several particulars. 

 The meshes are nearl}^ circular, being 2.5 mm to 3.5 mm across; their 

 beams are tolerably thick and composed of diactins, which are en- 

 tirely similar to those occurring in the gastral membrane. Many 

 previous authors, as Schmidt, Marshall, and Schulze, who discussed 

 some points of distinction between the internal diaphragms and that 

 of the terminal sieve plate in Aphrocallistes beafrix Gray, came vir- 

 tually to the same conclusion regarding this point. In the present 



