128 Transactions of the Society. 



pearing altogether ; the horizontal rays, though sometimes capitate, 

 more frequently extending into long sinuous whip-like filaments 

 (Fig. 14, Plate VII.), which often become branched, and thus give 

 rise to such forms as that of Fig. 8, Plate YII. The curvilinear 

 filaments of difierent spicules intertwine with one another in the 

 dermal surface, giving it a loosely woven texture, like a single layer 

 of cotton-wool filaments : in some cases they touch without uniting, 

 in others they are soldered together at the point of contact. 



The branched rays of Fig. 8 cannot be explained by supposing 

 secondary siliceous fibres to have been independently developed in 

 the dermal sarcode, and subsequently to have become united with 

 the spicular rays ; these branched rays can only be regarded as a 

 further development of such undulating forms as that of Fig. 12. 



Another form of sexradiate is shown in Fig. 18 : in this the 

 proximal ray has become excessively long, the horizontal arms 

 remaining comparatively short ; Figs. 2 and 4 are similar, but in 

 the latter, one horizontal ray is bent backwards in an elongate 

 S-like curve, and all its rays are capitate, except the proximal one, 

 which is sharply pointed. In Fig. 18 one of the horizontal arms 

 is suppressed, and in Fig. 2 the distal ray ; the number of rays 

 suppressed in difierent spicules is very variable, sometimes both 

 proximal and distal rays disappear, and only the horizontal arms 

 remain forming a simple cross. The microspining of the spicules on 

 the other hand is very constant, but the mode of termination just 

 the opposite, one, two, or three rays, or any number up to six 

 sometimes becoming capitate, the proximal ray, however, usually 

 remaining pointed. 



Some of the sexradiate spicules, those for instance with very 

 long proximal rays (Fig. 18) appear to accompany the bundles of 

 acerates which project beyond the dermal surface, their four 

 horizontal arms not being given ofi" in the dermis, but at some 

 distance outside it, after the manner of anchoring spicules. 



One cannot but feel some curiosity as to the function of these 

 various spicules, though without actual observation of the habits of 

 the living sponge it seems idle to speculate upon them. The 

 dermal spicules, however, evidently serve to support the dermal 

 membrane of the sponge ; the long acerates have probably, as 

 Bowerbank would maintain, a " defensive" action, and it certainly 

 seems just possible that both they and the projecting sexradiates, 

 especially the latter, may serve to capture and secure any minute 

 worms or other animals which in wandering over the sponge should 

 come in contact with their points. Nutritious material would be 

 freed from such animals at every puncture on becoming wounded, 

 and when subsequently decomposition set in, swarms of Bacteria 

 and other organisms would result, and a vast quantity of edible 

 material so be set free to be conveyed by the inhalent currents into 



