72 THEORY OP THE EARTH. 



In the first place, they indicate much more clear- 

 ly the nature of the revolutions to which they have 

 been subjected. The remains of shells certainly 

 indicate that the sea has once existed in the places 

 where these collections have been formed : But 

 the changes which have taken place in their spe- 

 cies, when rigorously inquired into, may possibly 

 have been occasioned by slight changes in the na- 

 ture of the fluid in which they were formed, or on- 

 ly in its temperature, and may even have arisen 

 from other accidental causes. We can never be 

 perfectly assured that certain species, and even 

 genera, inhabiting the bottom of the sea, and occu- 

 pying certain fixed spaces for a longer or shorter 

 time, may not have been driven away from these 

 by other species or genera. 



In regard to quadrupeds, on the contrary, every 

 thing is precise. The appearance of their bones 

 in strata, and stillmore of their entire carcasses, 

 clearly establishes that the bed in which they are 

 found must have been previously laid dry, or at 

 least that dry land must have existed in its imme- 

 diate neighbourhood. Their disappearance as 

 certainly announces that this stratum must have 

 been inundated, or that the dry land had ceased to 

 exist in that state. It is from them, therefore, that 

 we learn with perfect certainty the important fact 

 of the repeated irruptions of the sea upon the land, 

 which the extraneous fossils and other production^ 

 of marine origin could not of themselves have 



