SPONGES. 213 



the silica. At other times the silica is replaced by the peroxide 

 of iron, or, more rarely, by iron pyrites. On the other hand, a 

 few instances have been noticed where the original calcite has 

 been completely, or only partially replaced by siliceous ma- 

 terial. 



Another condition, in which the mineral substance of the 

 sponge has been entirely dissolved and removed, leaving the 

 empty moulds of the spicular mesh in the matrix, is not un- 

 common, but more frequent in specimens from Upper Silurian 

 than other deposits. 



As stated, the most common change in the mineral substance 

 of the sponge skeleton is the replacement of silica by calcite. 

 Examples of this replacement occur in all the great geological 

 divisions, but seem to be confined to calcareous and shaly de- 

 posits. The calcite which has taken the place of the silica is 

 crystalline, and probably infiltrated the moulds after the silice- 

 ous skeleton was dissolved and removed. When the spicules 

 were large or the spicular mesh not too closely interwoven, the 

 change has been effected without effacing the character of the 

 structure; but where the sponge skeleton was composed of 

 small and closely united spicules, not only the finer details of 

 the skeleton have been destroyed but in many instances the 

 entire internal structure is, if not obliterated, at any rate too 

 undefined to be satisfactorily determined. In no palaeozoic 

 sponge have I been able to distinguish the canals of the spi- 

 cules, and in but few instances their method of union, but 

 usually have had to rely upon a general idea of the spicules 

 and what the specimen preserved of the canal system. 



The undoubted siliceous sponges from the Lower Silurian 

 strata of northern Illinois are good examples of the replace- 

 ment by calcite. In these the spicular mesh retains sufficient of its 

 original structure to give us a fair idea of its details, while the 

 canal systems, though often very complicated, are clearly de- 

 fined and readily determinate. A number of the same species 

 occur at Beloit, Wisconsin, but here the specimens are so poorly 

 preserved that nothing whatever can be made out of their in- 

 ternal structure. I have also obtained several closely related 



