OP THE SPONGIADiE. 223 



now i-emains to be proved how they may be rendered 

 available in future descriptions of those animals. I cannot, 

 perhaps, better attain this end than by detailing the order 

 and mode of employing them in the description of species 

 contained in my own Manuscript ' History of the British 

 Sponges.' The following is the order in which these 

 characters have been taken for examination and descrip^ 

 tion : — 



1, Porm. 2. Mode of Growth. 3. Surface. 4. Oscula. 

 5. Pores. 6. Dermis, and Dermal Membrane and its 

 Spicula. 7. Skeleton and its Spicula. 8, Connecting 

 Spicula. 9. Defensive Spicula, — external, internal. 10. 

 Spicula of the Membranes, — tension spicula, retentive 

 spicula. 11. Sarcode and its Spicula. 12. Ovaria and 

 Gemmules and their Spicula. 



Colour. 



Habitat. 



Condition when examined. 



This order of description, or any other that the student 

 may prefer, should always be adhered to, and no part of 

 the specimen under examination that is present, and which 

 affords specific characters, should be omitted in the descrip- 

 tion ; so that, when no mention is made of particular organs 

 or classes of spicula, it may be presumed that they are not 

 present in the sponge in course of description. A certain 

 portion of these characters are always available. Thus the 

 skeleton, incurrent canals or cells, the sarcodous system, the 

 dermal and interstitial membranes, the pores, and the oscula 

 are always present, while the excurrent canals or the cloaca 

 are occasionally absent. The intermarginal cavities, if 

 present, are not always distinguishable, and the external 

 and internal defensive organs are, either one or both of 

 them, frequently absent. 



Specific characters should always be of a positive nature, 

 such as the presence and form of particular spicula or other 

 organs. It is a great mistake, in writing specific descrip- 

 tions, to make the differences between species to consist of 

 one or two striking essential characters only. Such a prac- 

 tice may answer tolerably well when there are but two or 



