BRITISH SPONGlADiE. 219 



close adherence of such portions to the dermal membrane 

 of the sponge would be very liable to lead to the belief that 

 they were really parts of that animal j but if the sponge be 

 carefuUy removed from the sinuous cavities of the shell and 

 washed previously to examination, I will venture to say that 

 no such erosive organs will ever be found embedded or 

 organically attached to the dermal membrane of the sponge. 

 I have often seen these cells in fragments of the membranes 

 of the oyster-shells, but have never been able to discover 

 any of them in conjunction with specimens of the sponge 

 taken from the perforated bouldered limestones from the 

 oyster beds of Tenby, or from the perforated surfaces of 

 the limestone rocks of St. Katherine's Cave, at the same 

 locality. In an oyster-shell from which I have thin sec- 

 tions, the perforations in its substance consist of nearly 

 straight tubes of about the same diameter throughout, and 

 each single one terminates in a round or oval cavity, varying 

 in its diameter from three to eight or nine times the dia- 

 meter of the tube ; up some of the tubes the sponge may 

 be seen slowly winning its way, with one or two spicula 

 disposed in a longitudinal direction, while the upper por- 

 tion of the tube and the large terminal cavity is completely 

 empty. In another part of the same specimen there is one 

 of these large, elon go-oval cavities, from which there are 

 four ccecoid canals radiating from it of different lengths, 

 two long and two short ; the cavity is entirely filled with 

 the sponge, apparently from a fifth canal connected with 

 the exterior of the shell. Now the condition of the sponge 

 in the four coecoid canals is relatively the same, that is to 

 say, all the parts in immediate proximity to the large cell 

 are filled with the sponge ; the two small ones entirely so, 

 the two long ones only partially so; the shortest of the 

 two long ones, for rather exceeding one third of its length, 

 and the longest for not more than one fourth of its length. 

 This condition of the sponge in relation to the canals and 

 cavities in the shell unmistakably indicate that the cavities 

 and canals were prepared previously, and that the sponge, 

 at the time that life was arrested, was slowly winning its 

 way into them. In the progression of the sponge up one 



