412 ON THE GEOLOGICAL RELATIONS 



Sometimes, the most delicate parts of the shell are per- 

 fect, even to the membrane of the hinge, while in others 

 they are compressed or broken, and even dissevered 

 and dispersed. The causes of all this are sufficiently 

 obvious. The animal matter of fishes is rarely pre- 

 served, as I already noticed; and, of the quadrupeds 

 and cetacea, only the bones are found, except in the 

 rare and peculiar cases elsewhere described. 



Though casts and impressions cannot be called 

 organic fossils, they are equivalent records. They 

 need not be here distinguished, being essentially the 

 same. In the case of leaves, there is often no other 

 record. In shells, the interior cast is a model. It is 

 often difficult to account for the disappearance of the 

 organic body in these cases ; as the stone must have been 

 previously indurated, when the cast is perfect, and 

 we do not see how that should remain when the 

 model was destroyed. In the calcareous rocks, where 

 the shell has combined with the stone, there is no 

 difficulty. And in this case also, the animal matter is 

 sometimes diffused through the rock, producing the 

 fetid limestones; a fact occurring too in the case of 

 fishes. The ammonites are sometimes filled with sand, 

 and the shell itself is silicified : yet when a shell is 

 filled with ftint, it remains calcareous, and is sometimes 

 crystallized, assuming at least a fibrous structure. In 

 the former also, the siphunculus sometimes remains 

 when the shell has disappeared. 



In the ligneous fossils, the wood is replaced by 

 sandstone, but the bark often remains, converted into 

 coal. Such are the well-known fossils of the coal strata. 

 In some cases, especially in the minuter fossils, the wood 

 is converted into charcoal. Leaves, or at least their bark, 

 have become coal: they are often, rather drawings than 

 even bas-reliefs of the originals. In other cases, as of- 

 ten happens in shale, the bituminous matter is diffused 



