MINUTE STRUCTURE OF STROMATOPOEA AND ITS ALLIES. 201 



this source of error, however, we have guarded by the personal 

 preparation of our slides from specimens which we had previously- 

 subjected to a careful examination by means of polished sections 

 and acids. 



(d) In the fourth place, there occur specimens (especially in 

 the Corniferous Limestone of North America) which unite in 

 themselves both the last-mentioned states of preservation, the 

 skeleton of the fossil in parts being calcareous, whilst the sarcode- 

 chambers are infiltrated with silica ; whereas in other parts the 

 skeleton is siliceous and the chambers are filled with calcite. 

 Specimens of this nature require no further notice here, as they 

 teach us nothing that cannot be learned from an examination of 

 specimens of either the second or third group. 



(e) In the fifth place, we find a few examples in which not only 

 is the skeleton siliceous, but the sarcode-chambers are likewise 

 filled with silica. In these cases, of course, thin sections consti- 

 tute the only mode of examination, as acids are wholly ineffective. 



As regards the general conclusion to be drawn from the various 

 classes of facts above enumerated, it appears to us to be unques- 

 tionable that the Stromatoporoids were originally calcareous in 

 their composition. This conclusion seems to us to be rendered 

 inevitable when we consider that specimens of the Stromatopo- 

 roids are almost invariably calcareous in beds in which the accom- 

 panying fossils are also generally unaltered, and that they are 

 only found to be siliceous in strata in which the other fossils (un- 

 doubtedly calcareous to begin with) are silicified — that in no case 

 known to us does microscopic examination show the skeleton to 

 be composed of distinct siliceous spicules or of separate sand- 

 grains — that many specimens are found in which the original 

 spaces of the fossil are filled with silica, whilst the skeleton itself 

 is more or less entirely calcareous — and that in many cases the 

 original calcareous skeleton has clearly been coated to its most 

 minute recesses with minute crystals of calc-spar before being 

 finally infiltrated with silica. If we were to believe that Stroma- 

 topora was originally and essentially siliceous, we should have to 

 believe that it had been, in innumerable instances, converted into 

 lime. The researches of Zittel ("Die Hexactinelliden," Abhandl. 

 der k. bayer. Akad. der AYiss. 1877) and of Sollas (" On Stauro- 

 nema'' Ann. & Mag. Nat. Hist. ser. 4, vol. xix. 1877) have, 

 indeed, shown that the Hexactinellid sponges are occasionally 

 more or less completely converted into carbonate of lime ; so that 



