402 Mr. II. J. Carter on Deep-sea 



is only 87- by 3-1800ths inch long, and the arms 29-1800tli8 

 inoli long respectively. Here also there is a tendeney to 

 bifurcation in tlie latter; while the large stellate (megastrellum) 

 of the interior, although of the same size as that of the fore- 

 going species, has for the most part only six arms, and these 

 are smooth ^ not microspined (tig. 47, d, h), as in the foregoing 

 species. 



Hah. Marine. 



Log. Above mentioned. 



Obs. With only a fragment of the cap])ing or petrous crust 

 of this form, this is all that can be stated descriptively about 

 its spicules ; and the general form of the entire sponge of course 

 is absent altogether. 



The specimen, however, is very interesting in a develop- 

 mental point of view ; for its spicules being in many instances 

 abnormal in form, especially the siliceous ball, shows how 

 intimately the latter is connected with a stellate, and how, in 

 all probability in its minutest form, it always originates in one. 

 Thus the siliceous ball, even when of full size, often presents 

 itself here in the form of a thick coarse stellate, with from five 

 to seven ai'ms, each of Avhich may present more or less of the 

 little stellate terminations which, in juxtaposition, make up 

 the tessellated surface of the matured and normally developed 

 ball, showing plainly that the latter belongs to the stellate 

 group of spicules. 



We see a similar development of the siliceous ball in Dr. 

 Bowerbank's illustrations of Geodia tuherosa (Proe. Zool. Soc. 

 1872, pi. 46. fig. 11) and especially in the abnormal develop- 

 ments given by Schmidt (Spong. Kiiste Algier, 1868, Taf. iv. 

 fig. 6) on the left side of the illustrations of Stelletta inter- 

 media^ where, as Stelletta has no siliceous balls, it is evidently 

 the abnormal development of the stellate itself, which closely 

 approaches that of the abnormally developed siliceous ball in 

 Geodia megastrella, var. lievisjnna. 



The fact, too, that the siliceous ball belongs to and probably 

 originates in a stellate form, bears upon the nature of the 

 stellates present in the embryos of Geodia megastreUa, which 

 altogether are respectively hardly larger than the full-sized 

 siliceous balls of this species, and therefore can only present 

 these balls in a rudimentary state — that is, in the stellate form. 



In the three species of Geodia above described, the fluked 

 anchoring-spicule somewhat diflfers in the form of its head, as 

 may be seen in the illustrations ; but this has not been insisted 

 on in the descriptions, because the form often ditFers so much 

 even in the same species. 



