DENDY— REPORT ON THE HEXACTINELLID SPONGES (TRIAXONIDA) 221 



The dermal skeleton consists, in the first place, of stout hexacts (figs. 20, 21, 22), 

 with the outwardly projecting ray much shortened and the inwardly projecting ray 

 lengthened. The rays are bluntly rounded at their extremities and usually quite smooth, 

 although a little roughening may sometimes be detected, especially towards the end 

 of the longest ray, which may also be more sharply pointed (fig. 21). 



Between the dermal hexacts occur numerous radially arranged sarulse of the form 

 shown in figs. 23, 24, 25. These remarkable spicules consist of a straight shaft, some- 

 times slightly roughened, and terminating at the inner end in a blunt point, while the 

 outer end is oval club-shaped and covered with stout, sharp, forwardly-directed spines 

 more or less fused together. 



The dermal skeleton is evidently very easily rubbed off" and is only well preserved in 

 places, especially in the membranous collars surrounding the polyps, where, as already 

 stated, the sarulse form a marginal fringe. 



The gastral skeleton consists of hexacts (figs. 26, 35, hex.) of the same general type 

 as the dermal hexacts, but the rays are more slender, generally less unequal in length 

 (though usually one still seems to be very short), and frequently knobbed at the 

 extremities. In the invasion of the original central cavity by the skeleton these spicules 

 become very irregularly arranged (fig. 35, hex.), but they may still be seen, crowded 

 together in the middle of the sponge, even after the central cavity has been completely 

 obliterated. They seem to retain their independence for a long time, but it is quite 

 possible that they ultimately become incorporated in the dictyonal framework. 



The parenchymal spicules are as follows : — 



(1) Very long and slender uncinates, sharply j)ointed at the two extremities and 

 with feebly developed teeth (fig. 32). , 



(2) Small spiny hexacts (fig. 27) lying between the trabeculse of the dictyonal 

 framework, with which their rays may become fused. (It is sometimes difficult to 

 distinguish these from new nodes formed by outgrowth and fusion of spiny processes 

 from the trabeculse (cf. fig. 36, pr.).) 



(3) Oxyhexasters (figs. 28, 29, 30). Very variable, with smooth, slender, sharp- 

 pointed rays. 



(4) Discohexasters (fig. 31). Rather compact, with not very long terminal rays, 

 about five to each principal. 



Although a good deal of the soft tissues of the sponge still remains, I have not been 

 able to make out the flagellate chambers. 



I do not think there can be any reasonable doubt that the " Sealark " specimen is 

 specifically identical with Topsent's type from the Cape Verde Islands, in spite of the 

 difference in locality. There do, it is true, appear to be some minor differences in 

 spiculation, as indicated by Topsent's account of the dermal hexacts and pentacts, 

 which, unfortunately, he does not figure, and by the absence of all mention by him of 

 the spiny hexacts (No. 2 above). It is probable, however, that these apparent difier- 

 ences would disappear if it were possible to make a direct comparison of the specimens, 

 and the agreement in other respects is so close that I feel justified in making an 

 identification. 



