BRITISH SPONGIAD^. 343 



as a means of discrimination to the almost total exclusion 

 of those of internal structure, in which may be found 

 striking and unfailing specific differences which never vary 

 under any circumstances of locality or modification of ex- 

 ternal form. 



Dr. Fleming has justly characterised this species as 

 being " massive, rising into short rounded branches ; the 

 fibres are coarser and the substance denser than those of 

 S.fiuviatilis ; the spicula, too, though similar in form, are 

 thicker and about one fourth shorter." This description 

 when both species are attainable, is good as regards the 

 differential characters : but, fortunately, there are essential 

 characters of much higher value, which exist in the spicula 

 of the dermal membrane and in those of the ovaria, neither 

 of which have, I believe, been noticed by previous writers 

 on these subjects. Those of the dermal membrane are, 

 under ordinary circumstances, very indistinct. If we ex- 

 amine the membrane in water between glasses, the spicula, 

 as they lie immersed in the sarcode, are scarcely to be de- 

 tected ; but if previously mounted in Canada balsam, they 

 become at once distinctly visible ; they are very numerous, 

 and are disposed over the membrane without any approxi- 

 mation to order and have an average length of ^nd inch, 

 and are ^rd inch in greatest diameter. They vary to 

 some extent in their dimensions, but their form is always 

 fusiformi-acerate ; the spines are abundant, conical, and 

 acutely terminated at all parts of the spiculum, but they 

 are not very strongly produced. The interstitial membranes 

 are also plentifully supplied with the same description of 

 tension spicula as those of the dermal membrane. The 

 dermal membrane of S.fiuviatilis is aspiculous, and in this 

 character, therefore, we possess an organic difference in the 

 structure of the parts, which leads us at once to a definite 

 and correct mode of determining the species, however 

 closely they may simulate each other in form. 



The skeleton spicula also differ in form from those of S. 



fiuviatilis ; in the latter they are purely acerate, that is, 



having the same diameter throughout the greater portion of 



the shaft of the spiculum, and attenuating only towards 



