J868.] DR. J. S. BOWEKBANK ON SPONGES. 123 



the way of the student, who would naturally conclude they were 

 organs always present in the sponge, while, in fact, he may search for 

 them in many dozen specimens of Ilalichondria panicea and other 

 allied species without once seeing a gemmule in situ. 



If we may judge of the extent of their distribution by their ascer- 

 tained presence in various genera, and by the negative evidence af- 

 forded by our knowledge of other descriptions of ovaria, it will not 

 be an overestimate if we believe that they will be found in at least 

 95 or 96 per cent, of the known siliceous and keratose sponges. 



Page 502, subsection 1. "Netted sponges (Z>«c^yo5/jo?7^2'«). Ske- 

 leton formed of a continuous siliceous or horny network." This de- 

 finition embraces so wide a range and such varied modes of structure 

 that it is calculated rather to bewilder the student than to facilitate 

 his researches. It would include rather more than 75 per cent, of 

 the whole number of the British sponges, and also the whole of the 

 author's second, third, and fourth orders, excepting his Tethyadce. 



Subsection 2, p. 503. The families included in this section, if 

 distributed in accordance with its wording, might just as correctly 

 be referred to subsection 1, page 5(;2, as to order 4, p. 504, sec- 

 tion 2. ChlamydosporeB is founded on a misapprehension of the 

 structure of the spicula of the ovaria, which have no defensive cha- 

 racters, but are simply portions of the wall of the ovarium, which is 

 so constructed that the spicula may allow of a slight degree of lateral 

 expansion of that body ; their external surfaces in the adult state 

 are flat coincident planes. The sponges having truly armed ovaria 

 or gem mules are entirely excluded by the author from this section, 

 the family Tethyadce being disposed of iu his order 4, Acantho- 

 spoiigia, p. 504. 



In truth, the confusion of ideas existing in the descriptions of 

 these orders is such as to render it perfectly impossible to compre- 

 hend the characters by which species are to be referred to any one or 

 the other of them. 



This course of proceeding is not divisional arrangement ; it is no 

 separation of the multitudes with the members of which we want 

 to cultivate an individual acquaintance. These vague expansive ideas 

 are not definitive or distinctive. What we want in such descriptions 

 is a certain group of characters within a definite circle to be readily 

 separated from other well-defined groups or circles by which they 

 are surrounded, such as we see amidst the natural orders of plants, 

 and not a flow of indefinite verbosity, that leaves us in a labyrinth of 

 words, amidst an impervious cloud of ideas. 



The Orders. 



The author's orders are seven in number. They are formed prin- 

 cipally by an elevation of his own and other writers' genera to that 

 dignified position in science, their characters being constructed by 

 a free version of their previous generic ones ; but in many instances 

 these alterations or additions have a very unfortunate result, as they 

 have led to serious errors of description. 



