160 Prof. W. J. Sollas on the 



each end of the spicule, up to which, indeed, it completely 

 reaches (PI. VII. "tig. 18). In these cases the spiculc-sheath 

 is no other than a single enveloping large cell ; and since the 

 spicule increases in thickness by successive onlayerings to its 

 surface, and nothing intervenes between it and the surround- 

 ing cell, we are obliged, so it would appear, to regard the 

 latter as the medium through which the spicule increases in 

 size ; but the very young spicules appear in a similar cell, 

 which only differs in being smaller and having a correspond- 

 ingly smaller nucleus and nucleolus, i. e. in being younger. 

 Hence it follows tliat the ensheathing cell is the true parent 

 and nurse of the large spicules with which it is associated ; 

 it probably only disappears on the completion of their 

 growth. 



The spicules most eons})icuously ]>resent are tlie projecting 

 forks, which, witli the grapnels with recurved rays and the 

 long fusiform acerates, lie in parallel bundles, radiating not 

 directly at right angles to the surface, but a little obliquely, 

 their inner ends being tangential to an imaginary sphere con- 

 centric with the centre of the sponge. The tendency to a 

 spiral arrangement, which Schmidt has well explained as 

 naturally following from the form of the spicules, is thus 

 early declared. Although the points of the forks frequently 

 project through the skin, yet their centre of origin from the 

 shaft always remains beneath or inside it. Most of the fully- 

 developed grapnels (and most of them are fully devcloj^ed, all 

 three rays being present, although O. Schmidt asserts they do 

 not become complete till the young sponge has left the parent) 

 are completely covered by the skin ; but those still incom- 

 plete, with only one or two teeth, as iigured by Schmidt and 

 Carter, more often protrude for some distance outside it, so 

 that they appear to be in a retarded stage of develoj^ment in 

 adaptation to some special requirements of the larval state — 

 it may be, for escape from the maternal tissues. 



There are no short fusiform acerates, such as occur in the 

 fibrous cortex of the adult ; but hamates in their various forms 

 are well represented. 



Distrihution. Kors Fiord, Norway, Station lo, 200-300 

 fathoms; Station IG, Station 23, 180 fathoms. 



The species is also found about the Shetland Islands in 

 deep water; Iceland, Florida : 152-183 fathoms. 

 The other described species of the genus are : — 

 1. T. antarctica, Carter, A. & M. N. H. 1872, vol. ix. 

 p. 412, pi. XX. Loc. Antarctic Ocean, lat. 74^° to 77^° S., 

 long. 175"^ W. ; depth 206-300 fms. Distinguished by ab- 

 sence of hamates. 



