BKITISH SPONGIADjE. 123 



and the hispid character scarcely perceptible, but in old and 

 well-worn specimens it is frequently rugose, and more or 

 less excavated. The oscula are not readily detected in the 

 fresh sponges ; when apparent they appear as sim-ple circular 

 orifices, rather numerous, minute, and irregularly dispersed 

 on all parts of the sponge. The dermal and interstitial 

 membranes appear entirely destitute of tension spicula ; the 

 former tissue is perforated by the distal terminations of the 

 secondary ramuli, which form innumerable little radiating 

 groups of defensive spicula, projecting through the mem- 

 brane to the extent of about two thirds the length of a 

 spiculum. 



In the dried condition the primary cylindrical radii of 

 the skeleton have very much the appearance of semi- 

 decomposed ligneous fibre, and the exposure of the secon- 

 dary ramuli by the contraction arising from drying gives 

 the surfaces of the sponge a remarkably flocculent appear- 

 ance, like that of a close thick coat of fine hair or fur, laid 

 in every possible direction by immersion in water. 



When sections of this sponge in a longitudinal direction 

 are mounted in Canada balsam, and examined by trans- 

 mitted Ught, with a power of about 150 linear, the primary 

 ramifications of the skeleton are seen to consist of in- 

 numerable long and very flexuous spicula, loosely fascicu- 

 lated together, as in the corresponding parts of a Dictyo- 

 cyhndrus, but the spicula in this sponge are very much more 

 contorted in their flexuosity than in any species of Dictyo- 

 cylindrus with which I am acquainted ; and on the external 

 parts of the cylinders these spicula are frequently so 

 dispersed as to appear as if they were portions of threads 

 which had been bound round the central' fasciculus in an 

 ascending direction, to keep the whole of the included 

 spioula in position. 



The secondary ramuli of the skeleton are singular in 

 their structure. Their proximal terminations appear to 

 have scarcely any connection with the primary cylinder 

 whence they emanate; they commence with one, two, or 

 three spicula loosely cemented together, but they accumu- 

 late spicula rapidly as they increase in length, dividing 



