236 STRUCTURE AND 



points, in the process of formation, (I had almost 

 said, of crystallization J have shot through and among 

 each other, so that it would he almost impossible to 

 extract one from any point without either breaking 

 off its rays, or tearing away a considerable portion of 

 the whole surface. The rays shoot in the same plane, 

 and in that plane the stars lie, not quite at random, 

 as to their direction ; for the great majority have one 

 point directed lengthwise from the mouth of the tube 

 towards the base. There are not wanting however 

 many, which point in the opposite direction ; and 

 several at intermediate angles. Of course, it requires 

 but little divergence from the first named direction to 

 produce the second ; still, however, the prevalent 

 order appears to be this. 



I cannot trace any fibrous or gelatinous or granular 

 matter in which the spiculee are set; but beneath the 

 layer formed by their interlacing points, there is a 

 surface composed of round granules of transparent or 

 pellucid matter, set as close as possible, which are 

 plainly seen between the crossing needles. This ap- 

 pears to be the interior lining of the tube, in fact the 

 tube itself, around which the spiculse are arranged as 

 a loose outer casing, giving firmness to the whole. I 

 could not detect spiculse of any other form than the 

 three-rayed stars ; but several of these had one or 

 more of their rays broken short; for from their com- 

 position they are very brittle, as I have often proved 

 in other species. 



The form of this specimen was so very irregular 

 that but a poor idea can be conveyed of it by words : 

 it may, however, be roughly described as an elliptical 



