402 W. J, Sollas — A New Cretaceous Sponge. 



and now and then I have seen a small solid central sphere inside, 

 looking like a nucleus. In size they vary from a+o-" to toVo-" 

 (PL XIV. Fig. 6) ; they are often amber in colour when viewed with 

 transmitted light, and they appear to be soluble in acid, though it is 

 difficult to make this out with certainty in the case of such very 

 small bodies imbedded in a chalky matrix, which is itself soluble 

 with effervescence in the same solvent. Possibly they are single 

 detached chambers of Foraminifera. like Glohigerina or Nodosaria, 

 both of which occur in the sections along with Greensand casts ; but 

 the abundance of these round bodies, out of all proportion to the 

 number of the Foraminifera associated with them, their perfect 

 shape, and complete isolation, would seem to negative such an idea, 

 and half incline one to look upon them as independent organisms. 

 I mention them here because both they and the contort branching 

 bodies before referred to occur very commonly in the ordinary 

 so-called '' Coprolites " of the Cambridge bed. 



As regards the position o{ Euhrochus in our classification, it clearly 

 belongs to the HexactinelUdce, since in no other sponges do we find 

 the same rigid adherence to a sexradiate type. The existence of 

 hollow casts only in the place of the original skeleton might appear 

 at first sight difficult to reconcile with the view that the latter was 

 composed originally of siliceous material ; but here we are no worse 

 off than in the case of the Ventriculites preserved in a phosphatic 

 matrix, in which there has occurred a similar disappearance of fibres, 

 of whose original siliceous nature there can now be no manner of 

 doubt, since their characteristic structure and arrangement have been 

 discovered, in all their details, in the vitreo-hexactinellid Myliusia 

 Grayi. 



Having then placed our sponge amongst the HexactinelUdce, we 

 must assign it to that family of the order in which the skeletal 

 spicules are united together into a rigid continuous network [Vitreo- 

 Jiexactinellidid), instead of being merely connected together by sarcode 

 [Sarco-JiexactinelUdce) . 



A closer determination of the alliance of Fubrochus is very difii- 

 cult to make, especially as the rosettes, and other forms of flesh 

 spicules which characterize the HexactinelUdce, are entirely absent ; 

 that they once existed may be regarded as highly probable, and their 

 disappearance can excite no surprise, since the processes of solution 

 which have destroyed the coarser fibres of the skeleton would inevit- 

 ably remove every trace of these more delicate structures. The 

 dermal reticulation might be expected to furnish some clue, and, 

 indeed, it closely resembles in the squareness of its meshes the 

 single layer of lattice-like network which constitutes the skeleton 

 of Farrea occa. Fubrochus differs widely from Farrea, however, in 

 the fact that its external reticulation forms a complete in closure to 

 an interior skeleton, while in Farrea a single layer of harrow-like 

 network simply forms an empty tube open at the mouth. The resem- 

 blance with Farrea, then, is solely in the rectangular meshwork of the 

 dermal reticulation, and the value of this for classificatory pur- 

 poses is very doubtful, since Dr. Bowerbank has recently described 



