146 THE GEOLOGICAL RECORD. 



or we must allow the various fossil species to be con- 

 nected with one another by descent. Upon the latter 

 alternative, a whole chain of transitional links must once 

 have existed between the earliest form of oyster and the 

 latest ; and though many of these links have been pre- 

 served,, still more must have been lost, or deprived of 

 their distinctive features ; so that here, where the geo- 

 logical record is, to all appearance, unusually perfect, its 

 actual imperfection is more clearly than ever established. 



To conclude, then, in few words : The majority of 

 dead creatures never become fossils at all. 



The majority of fossils perish miserably in their 

 hiding-places. 



Of those that are saved, the majority cannot be got 

 at by man. 



Of thdse that can be, the majority never are. 



Of those that are, a large number prove illegible ; a 

 large number are fragments ; a large number duplicates ; 

 and, lastly, a large number fall into hands which again 

 lose or destroy them. 



We cannot therefore argue, because fossils of such- 

 and-such forms of life have never been found, that such- 

 and-such forms never existed. They may have existed, 

 and left no fossils. The fossils may have been left, and 

 subsequently destroyed. The fossils may be undestroyed, 

 but never have been found. The sum-total of acquisi- 

 tions is small, but precious ; it never can make a com- 

 plete record, but it may make one sufficiently ample to 

 establish the Darwinian Theory, or to replace it by some 

 still wider and still simpler generalization. 



