BRITISH SPONGIADiE. 47 



they are the intermarginal cavities which receive the minute 

 streams from numerous pores situated immediately above 

 and within a short distance of them ; the true pores per- 

 forating the dermal membrane are too minute to be visible 

 without the assistance of considerable microscopic power. 

 In the dried specimens these " dimpled" depressions of 

 Dr. Johnston, or stellate orifices of Dr. Fleming, have the 

 appearance of forming a well marked character of the sur- 

 face, while, in truth, I believe they are not visible on the 

 surface of the sponge in its natural condition before it is 

 dried ; at least this is the case with G. Barreltii. In the 

 dried specimen they are apparent, but not in the slightest 

 degree visible in any part of the same specimen that was pre- 

 served in salt and water while fresh from the sea ; we may 

 therefore reasonably conclude, that they do not form a 

 character applicable to the description of the species. 



In the present condition of the sponge the villous cha- 

 racter described by Dr. Fleming is not in the slightest 

 degree apparent, excepting in the depressed parts before 

 mentioned, but if we examine a section of the sponge at 

 right angles to its surface, we find that remains of large 

 spicula passing entirely through the crustular dermis are 

 abundant ; they originate at a short distance beneath the 

 inner surface, pass through the crustular mass, and project 

 at least half their length beyond its outer surface. In its 

 natural condition it would therefore be decidedly hirsute. 

 There are also a few small acerate spicula at right angles to 

 the surface, but these rarely project beyond the dermal 

 membrane, and are not sufficiently long to reach the inner 

 surface of the crust. 



Where there are any remains of the dermal membrane it 

 is profusely furnished with attenuato- stellate spicula, the 

 same as those of the sarcode, but its dilapidated condition 

 afforded very unsatisfactory information regarding the pores; 

 apparently they are equally dispersed over its surface. 



The connecting spicula are of two distinct forms, and are 

 very large and long. The attenuato-expando-ternate ones 

 are very stout and strong ; the triradiate heads of this form 

 are generally parallel to each other at the base of the 



