122 DR. J. S. BOWERBANK ON THE SPONGIAD.E. [Feb. 6, 



tlier, and remain so until they reach the point of their natural diver- 

 gence immediately beneath the dermal surface, at which point they 

 are usually closely broken off. 



One of the most interesting subjects in the history of this little 

 species is the mode of its propagation, which appears regularly to be 

 by an abundant production of external gemmules. 



Of the thirty-six specimens in my possession, twenty-two have 

 more or less of gemmules attached to their surfaces. The largest 

 uumber on any one sponge is eleven, on the little specimen re- 

 presented by fig. 1 8, Plate V. 



They are projected from all parts of the surface. The length of 

 their fleshy thread-like attachments to the parent sponge varied 

 considerably ; and in one gemmule very perfectly developed, with a 

 diameter of 1| line, the attaching filament was \ inch in length 

 and very slender. On many of the mature specimens of sponges 

 on which there were no gemmules remaining, there were short 

 fleshy filaments which had every appearance of being the remains 

 of the attachments of these little bodies which had separated from 

 the parent sponge. 



The attachments of the mature specimens are various, in accord- 

 once with their necessities. Sometimes it is effected on smooth 

 surfaces by a close adherence of the basal portion of the sponge, 

 while in other cases they project short root-like appendages varying 

 from one to four or five in number, the distal extremities of which 

 frequently expand to a considerable extent over the rough surfaces 

 of dead shells or other substances on which they have located them- 

 selves ; and in one case of a Tethea thus located on a dead shell 

 and having gemmules produced near its base, these little bodies, 

 still attached to the parent sponge, had themselves projected small 

 cords of attachment from the parts of their surfaces nearest the 

 shell surface with the evident intention of securing a permanent 

 attachment previously to separation from the parent. 



The tension spicula of the interstitial membranes are very long 

 and slender ; they are mostly dispersed in lines more or less accord- 

 ing with those of the skeleton-fasciculi. 



> 



Genus Halispongia, Bowerbank. 



De Blainville, in his ' Manuel d'Actiuologie,' p. 532, proposed the 

 name of Halispongia to receive all those sponges that Fleming and 

 Grant had designated Halichondria. But as our recent aud more 

 extensive knowledge of British and exotic species comprised under 

 the names of Halichondria and Halispongia has demonstrated their 

 widely different skeleton-structures, it has become necessary to divide 

 this very extensive group into a series of new genera in accordance 

 with their organic structural affinities. I therefore proposed, in 

 vol. i. p. 207 of ' Monograph of the British Spongiadae,' to limit the 

 genus Halispongia to such sponges among the kerato-fibrous suborder 

 as agree with the following character : — 



Skeleton kerato-fibrous. Fibres solid ; primary fibres compressed, 



