8 G.H.Parker 



epithelial cells, as in fact they are believed to be in many other 

 sponges (Minchin, 'oo, p. 46; Schneider, '02, p. 260). Even admit- 

 ting, hov^ever, that in Stylotella they are in all cases covered by an 

 epithelium, the epithelium is certainly in most instances so 

 extremely thin that these cells are almost in contact with the sea- 

 water passing through the cavities that they surround. 



In the region of the osculum the myocytes are especially numer- 

 ous and form a conspicuous sphincter on the inner face of the 

 oscular collaT and internal to the mass of longitudinally arranged 

 spicules which surround this opening. As a result the contraction 

 of the osculum is accomplished without much folding of the surface 



Fig. 3. Transverse section of the base of an oscular collar of Stylotella. The central cavity is the 

 osculum, which, as is shown on the right is directly surrounded by a sphincter of myocytes, external 

 to which is the tissue containing the spicules. X 35. 



of the sponge next the oscular cavity, for this surface is immedi- 

 ately in contact with the contractile material, if in fact it is not 

 contractile itself. On the other hand the outer substance of the 

 oscular membrane and the palisade of spicules are thrown into 

 many folds in contraction as though they passively followed the 

 constricting ring of myocytes to the outside of which they are 

 attached (Fig. 3). This palisade-like arrangement of the rigid 

 siliceous spicules is the only one that would allow an easy con- 

 traction and expansion of the osculum and it is in strong contrast 

 with that of the spicules in the rest of the sponge, in which these 

 bodies show no such grouping. 



