PLATE LXXXVII. 305 



I received this sponge preserved in spirit from my 

 friend the Rev. A. M. Norman, who dredged it at 

 Shetland in 1867. It covers nearly the whole of the 

 surface of the fragment of a bivalve shell, and it does 

 not exceed about the fourth part of a line in thick- 

 ness. 



At the first examination of a piece of this sponge 

 mounted in Canada balsam with a power of about 150 

 linear the spicula appear to be confusedly mixed 

 together, and it is not until after a close and careful 



^^ 



examination with a still higher power that the appa- 

 rent confusion becomes comprehensible and the mixed 

 mass of spicula can be assigned to their respective 

 offices, and the abundance of the dark amber-coloured 

 sarcode in which they are embedded increases the 

 apparent confusion. 



The tension spicula of the dermal membrane are 

 purely acuate in their form, the diameter being equal 

 throughout the whole of the length until very near 

 each termination, when they are rapidly contracted 

 and terminate acutely. A few of them are dispersed 

 singly, but their mode of disposition is usually in 

 loosely formed fasciculi, which appear either isolated 

 or crossing each other irregularly. 



Their mode "of disposition is not very readily dis- 

 tinguishable, even when a portion of the sponge has 

 been mounted in Canada balsam, in consequence of the 

 abundance of deeply coloured sarcode, and of the 

 numerous and confused mass of spicula immediately 

 beneath them. Under these circumstances they 

 require a power of about 250 linear to render them 

 distinctly visible. The two forms of retentive spicula 

 in the dermal membrane, are very minute and delicately 

 formed, they are few in number, and it is rarely that 

 they can be detected in situ. 



The skeleton fasciculi are loosely constructed, and 

 very irregularly disposed, and interspersed with them, 

 there is a considerable number of single spicula of the 

 same form, some of which are twice the diameter and 



20 



