132 Mr. H. J. Carter on Specimens 



cause it points out the relation and function of this layer, 

 which, in most specimens, is so absorbed into the petrous 

 crust as to be undistinguishable from it ; while the futility of 

 attempting to circumscribe the uses of any thing in the 

 Spongida is also shown by the head of the zone-spicule, 

 generally considered to be for supporting the petrous crust, 

 here passing beyond it so as to support alone the dermal 

 sarcode. Of course, all that passes in through the pores must 

 go into this subdermal interval or chamber previous to passing 

 into the body of the sponge ; moreover there are the same 

 hourglass-shaped holes in the petrous crust as those which I 

 so minutely described and illustrated several years ago in 

 Pachymatisma Johnstonia, Bk. ('Annals,' 1869, vol. iv. p. 12, 

 pi. ii. figs. 11, 12). It is not my object here to trace the 

 progress of this material in through the petrous crust ; but I 

 cannot help thinking that the larger holes or vents in the 

 dermal layer are, in all probability, the openings of the ter- 

 minal canals of the excretory system, which pass through the 

 chamber entire, so that their contents may not be mixed with 

 the incoming ones through the pores. 



This kind of dermal layer 1 described and figured many 

 years ago in Svongilla under the name of " investing mem- 

 brane " ('Annals,' 1857, vol.'xx. p. 24, pi. i. fig. 1,111), 

 showing that it is the seat of the dermal pores, while it is 

 supported on the ends of spicules which, projecting from the 

 parenchyma of the sponge, form a hollow space between 

 it and the latter, into which the water and its contents are 

 received before passing further into the interior. Moreover 

 it is shown that the terminal canal of the excretory system is 

 continued through this space or chamber entire, so that its 

 contents may not mix with those of the chamber (ib. ib. 

 fig. l,g), which I have assumed to be the case in Geodia per- 

 armata. And lately, again, I have represented it in a marine 

 sponge, viz. Halichondria simulans ('Annals,' 1874, vol. xiv. 

 pi. xxii, fig. 34, e e) ; while Dr. J. Millar has called atten- 

 tion to its existence over the Hexactinellida, where it is sup- 

 ported in a similar way, strengthened by a square lattice- 

 work of hexagonal flesh-spicules, just as in Geodia perar- 

 mata it is supported on the heads of the zone-spicules and 

 strengthened by the addition of stellate flesh-spicules. 



Although these parts are not all illustrated in Dr. Bow- 

 erbank's description and illustrations of his Geodia per- 

 armata (Proc. Zool. Soc. Jan. 7, 1873, p. 8, pi. ii. figs. 1 

 &c), still the spiculation there given is so like that of the 

 Manaar specimens that, taking into consideration his statement 

 (p. 9), that the heads of the zone-spicules are " projected 



