70 THE LESSON OF EVOLUTION 



are inaccessible to us. We can only examine the 

 outcrops in cliffs or quarries, while the rest is buried 

 too deep for excavation. We must, therefore, put 

 limited trust in negative evidence, and cannot say 

 that an animal or plant did not exist at a certain time 

 because we have not found its remains. 



But, on the other hand, we have already collected 

 an enormous number of fossils from nearly every 

 part of the world, and each of these fossils forms 

 positive evidence which cannot be gainsaid. More 

 than that : this mass of positive evidence is suffi- 

 cient to enable us to make inferences bearing on the 

 negative evidence, and so often enabling us to judge 

 how far we can trust to it. For example when we 

 find numerous leaves of dicotyledonous plants with 

 comparatively few ferns, in the rock younger than 

 the Jurassic, and abundant ferns without any dico- 

 tyledonous leaves in the rocks older than the 

 Cretaceous and younger that the Silurian, we may 

 feel sure that dicotyledons either did not exist or 

 were very rare in the older periods. It is the same 

 with the terrestrial mammals, the bones of which 

 we find in the Cainozoic rocks, where terrestrial 

 reptiles are hardly represented, an exact opposite 

 to the rocks of the Mesozoic Era. Here also we may 

 make a similar inference. 



Again, the order in. which animals and plants 

 appeared on the earth is a remarkable one, because 

 it coincides with the arrangement arrived at in the 

 classification of living plants and animals, and this 

 is evidently not a chance. This shews that we can 

 trust the palasontological evidence, when viewed on 



