Sponges jrom the Atlantic Ocean, 323 



form ones may be considered as " incomplete development." 

 (If there is one thing more to be deprecated than another in 

 the description of sponges, it is the figuring of exceptional 

 forms of spicules as characteristic of the species.) 



There is, however, a great diversity of form in all three 

 kinds of spicules, since the terminal inflation of the large 

 spicule is not only occasionally double, and that of the cen- 

 trally inflated spicule also, but the extremities of the latter, 

 although always more or less fissurate or spined, are equally 

 varied. 



Then, again, the verticillately spined and moniliform spicules 

 vary in size from 2- to 20-1800ths inch in length, while the 

 absence of any particular form of flesh-spicule may be sup- 

 plied by the smallest verticillate ones, in which the central 

 inflation then causes them very much to resemble the centrally 

 inflated flesh-spicule of Halichondria suherea and H. Jlcus, 

 Johnst., Suherites domuncula, Sdt. (Dr. Bowerbank, op. cit. 

 vol. ii. p. 202, is wrong in restricting the presence of these 

 centrally inflated flesh-spicules to H. ficus^ inasmuch as they 

 are equally present in both the type specimens of H. suherea 

 and H. jicus respectively, of the Johnstonian collection in the 

 British Museum.) 



The only approach in form to the centrally inflated subske- 

 leton-spicule with fissurate ends of HymerapJiia verticillata^ 

 that I know of, is in Halicnema patera^ Bk. (vol. iii. pi. xv. 

 figs. 31 and 32); but here the ends are sharp -pointed, although 

 the centre of the shaft is once and sometimes twice inflated ; 

 still these spicules are congregated round the great sub-pinlike 

 acuates of the fringe at the circumference of H. patera^ where 

 they thus bear the same relation to each other that the cen- 

 trally inflated spicules do to the great sub-pinlike spicule in 

 Hymeraphia verticillata. The double terminal inflation of the 

 latter, too, is common in Halicnema patera^ while the staple 

 spicule of the body generally, which is smaller, consists of a 

 curved acerate, inflated in the centre, and thickly (although 

 not verticillately as in KymerapMa verticillata) spined through- 

 out. So that the spicule-complement of Halicnema patera 

 comes nearest of all known sponges to that of Hymeraphia 

 verticillata ; and the former I have thought best for the present 

 to place among the Suberitida. Perhaps Halicnema patera 

 and its like may have to come there also. 



It has been above stated, conjecturally, that the great sub- 

 pinlike-spicule which projects from the summit of the aculea- 

 tion is about 200-1 SOOths inch long (that is, -g- inch) ; but as 

 this spicule from its extreme length is generally broken off 

 just outside the summit of each aculeation, while its inner 



