REVISIOX OF FRESHWATER SPONGES OF SPONGILLIDAE 87 



and rather ill-defined horizontal spicule fibers, joined together by a 

 considerable amount of spongin. Consistency of live sponge firm and 

 moderately hard, skeleton of dry sponge compact. Bubble cells 

 abundant in inner sym plasm. 



Megascleres straight to sUghtly cm-ved and stout amphioxea, 

 distinctly fusiform, typically armed with small but acute spines except 

 at their tips, very rarely incipiently spined or altogether smooth; 

 length range 200-350 n, mdth range 9-20 /x. 



Microscleres absent. 



Gemmoscleres birotulates of one class, with a moderately thick and 

 short shaft, as a rule smooth, and terminally with rotides of equal 

 diameter and distinctly flat shape, irregularly and deeply incised into 

 a smaller number of long rays, usually not more than 12; maKorma- 

 tions frequent in adverse habitats; length of shaft typically 12-20 n, 

 diameter of rotules ranging 20-25 ju, width of shaft 4-6 m- 



Gemmules moderately abundant in mature sponge, either scattered 

 in the skeletal mesh work, or aggregated in the basal region \\-ithout 

 forming a pavement layer; they are spherical to subspherical in 

 shape, ranging in diameter 350-450 n; pneumatic layer well developed 

 but shallow, consisting of minute subspherical air spaces; gemmoscleres 

 embedded in this coat typically in one layer, occasionally forming an 

 additional distal layer on top of the first; outer gemmular membrane 

 usually feebly developed or altogether lacking, distal rotules of 

 gemmoscleres clearly perceptible; foramen only very slightly elevated, 

 never tubular. 



Distribution. — ^Apparently restricted to the Northern Hemisphere, 

 mth preference to cold- to warm-temperate regions, widely distributed. 



Color in life. — Usually drab yeUow to brown, occasionally green 

 due to presence of zoochlorellae. 



Discussion. — This species has been adequately described in 

 pre\'ious literature and the present studies have not revealed any 

 additional criteria for discussion. As can be expected in sponges with a 

 \videly scattered distribution, E. mulleri displays some slight variations 

 in distant populations which, however, are not significant enough to 

 permit a clear discrimination of races. The multiple arrangement of 

 gemmoscleres, as sometimes present, as well as the occasional mal- 

 formation of megascleres certaiiJy cannot be used for such a purpose. 

 However, the condition displayed by E.japonica, in spite of the various 

 argimaents involved in this problem, is here considered to represent at 

 least a racial variant of importance. In order not to obscure perceptible 

 taxonomic evidence, E. japonica in this paper is dealt with as a sepa- 

 rate species until evidence to the contrary can be obtained by future 

 detailed studies. 



