76 THE SPONGES. 



waterworn, it is by no means certain that they represent the upper end 

 of the perfect sponge. There is, however, a clear indication that the 

 sponge bifurcates above. 



The lateral edges of the lenticular, 'or flattened, stem are everywhere 

 rounded, nowhere sharp. At the Tipper end, however, in the above- 

 mentioned notch, the free edge becomes thin and comparatively sharp. 

 The stem is slightly curved in a wave-like fashion both along its edges 

 and its flattened faces, but very vaguely and in a less regular fashion than 

 in Schulze's specimens of B. subtilis. The wave-like contours nevertheless 

 suggest that the growth is a bilateral modification of the spiral form. 

 It may here be mentioned that a strongly bilateral modification of the 

 spiral form is exhibited by some of the macerated specimens, which I 

 mention under Eurete erectum, p. 58. In one of these, over a tract 70 mm. 

 long, the curves of the stem lie nearly in one plane, and the lateral cups 

 form two linear series, which are distributed along the opposite curved 

 edges of the sponge, the curved edge presenting a wave-like contour from 

 the convexities of which arise the cups. 



The dictyonal framework has the structure described by Schulze. 

 In the middle of the comj^ressed stem there, are longitudinal beams 

 with more or less transverse connectives, the system thus giving fairly 

 rectangular meshes. The longitudinal beams on each side curve outward 

 toward the edge of that side and the upper end, in some cases obviously 

 branching acutely. Connectives extend transversely between these beams 

 and at about right angles to them, and thus make an angle with that 

 horizontal axis of the sponge which runs from edge to edge of the stem. 

 Typically the connectives form continuous lines which extend from edge to 

 edge of the stem, and are strongly arched toward the apex of the sponge, 

 precisely as described by Schulze for B. suhtiiis. There are of course 

 departures from this plan, owing to the fact that some connectives lie 

 in the transverse axis of the stem, others extend outward and upioard, 

 and at some levels the connectives do not form continuous lines. 



Away from the median plane, going toward each face of the stem, the 

 skeletal meshwork becomes irregular. Near and at the surface, the beams 

 are much slenderer and the meshes larger than farther in. 



Abundant small hexacts are present. It should also be mentioned that 

 a few hexact pinuli and scopulae were observed. Since these, however, 

 resemble the corresponding spicules of Eurete erectum gradle, specimens of 



