12 ART. 9. 1. IJIMA I HEXACTINELLIDA. 



The type of G. elegans is a complete specimen ; only it is 

 strongly macerated on the external surface, so that the cuff and 

 parietal ledges, if these were originally at all present, must have 

 entirely fallen away. A considerable quantity of the loose tissue 

 however still exists in connection with the skeletal lattice-work, 

 while the internal surface of the wall seems to be tolerably well 

 preserved. Nothing unusual in the quantity of the loose tissue 

 on this surface attracted my attention ; it is of much the same 

 character and appearance as I know it to be in Regadrella or 

 in Euplectella. The statement of Marshall ('75) that the 

 " flake-tissue " lies on the inner side of the skeletal lattice-work 

 applies to the actual condition only in a relative sense that that 

 tissue has been lost from the outside by abrasion. 



The clavate tubular body measures 232 mm. in height, In- 

 feriorly it gradually narrows towards the solid basal knob, close 

 to which extremity the width is 34 mm. The irregularly rounded 

 upper extremity is 101 mm. wide, as measured in one way. The 

 wall at this end shows a considerable — evidently an abnormal — 

 outbulging towards one side, so much so that the body appears 

 as if bent at the top. For the measurements given above I am 

 indebted to the kindness of Professor Max Weber. 



Of the main parenchymal bundles or the skeletal beams, not 

 a few pursue a more or less distinctly longitudinal course ; the 

 rest run more or less obliquely, freely intersecting and anasto- 

 mosing with one another. Thus the skeletal lattice-work is of 

 an irregular character. A rather extensive area at the upper 

 end is to be considered as representing the sieve-plate. The 

 beams and the meshes at that part look much like those in the 

 sieve-plate of either Regadrella okinoseana or Euplectella mar- 

 shalli. However, as was mentioned by Marshall, they pass over 



