Potash on the Siliceous Skeletons of Sponges. 295 



film separated itself from the interior mass of a globo-stellate, 

 which everywhere remained free within it. The film was not 

 smooth, but wrinkled all over so as to fit exactly on to the 

 spines it once covered, dipping into the depressions between 

 them and rising over their summits (PI. IX. fig. 14, /) . Is this 

 film the silicified wall of the cell in which the spicule has been 

 developed ? or is it the last coating of silica which the spicule 

 had received? If the latter be the tnie interpretation, the 

 phenomenon is but a repetition of what we have already ob- 

 served in the case of elongated spicules, where the outermost 

 lamella frequently remains after solution as a mere sheath 

 about the rest of tlie spicule within it. 



In this sponge the potash is more frequently able to attack 

 the component rays of the globo-stellates from the exterior than 

 in Geociia arabica, and hence the solution in a radial direction 

 sometimes proceeds both from within outwards and vice versa 

 simultaneously. 



The stellates of Pachymatisma appear to have greater power 

 of resistance than those of Geodia arabica^ since a slide show- 

 ing many of the globo-stellates of the former reduced to the 

 condition of thin shells also contains instances of its stellates 

 which have survived solution, and, indeed, appear but little 

 the worse for it. AVith this fact may be coupled another, viz. 

 that stellates similar to those of Pachymatisma have been seen 

 by me in tlie fossil state amongst a number of other spicides 

 which I am now describing from the Chalk of Trimmingham, 

 Norfolk. 



Trachya^ sp. 



The staple spicule here is a large acuate, the rounded end 

 of which affords great protection to the corresponding termi- 

 nation of the axial canal within it, while at the pointed end 

 only a thin layer of silica intervenes between the axial canal 

 and the surrounding potash. It hence happens that at the 

 rounded end the canal remains closed and the concentric la- 

 mella) of the spicule are there with difficulty dissolved away; 

 while at the pointed extremity the canal is entered by the 

 potash at once, the edges of the concentric lamina are exposed 

 and rapidly attacked, and the spicule is eaten back from the 

 point towards the butt, so that it is soon nearly all destroyed, 

 the part that remains often being only the butt-end, which 

 even then preserves the blind termination of the canal un- 

 breached from behind (PI. IX. fig. 19,^). 



At the round end the potash removes the concentric lamella} 

 from without inwards, so that the innermost project furthest 

 (PI. IX. fig. 19, i) ; at the pointed end the potash acts from 



21* 



