210 Geological Evidence of Origin of Species. 



of birds from reptiles, or viceversa; but I contend that they 

 prodigiously lessen the improbability of such development ; 

 they supply us, at least, with some of the steps by which 

 the change may have taken place. Now we must bear in 

 mind the extremely fragmentary character of the geological 

 record, that our systems of strata are but a few irregular 

 rungs of the ladder reaching from the abyss of the past to 

 the recent period; that of these fragmentary systems two- 

 thirds are under the sea, and a portion of the remainder in 

 inaccessible polar regions ; that of the accessible areas of the 

 earth's crust not one tenth has been explored, and probably 

 not one hundredth carefully investigated, and that of this 

 small remainder but a small fraction is exposed in quarries 

 and natural sections. If, under such limitations and in so 

 short a time, so much has been done to bridge over the gaps 

 between different types, how vastly would our material be 

 increased if we could summon before us the long races of 

 beings who have left for our examination such fragmentary 

 relics of their existence ! Let me give an illustration. A 

 few genera of mammals were known from strata of Mesozoic 

 age. Suddenly, the exposure of a few square feet of Pur- 

 beck Oolite increased our knowledge threefold. Again, 

 between the Purbeck with its mammals and the Eocene 

 with its abundant mammalian fauna is the great series of 

 the Cretaceous system, in which no trace of a mammal has 

 been found. But would it be fair to infer that mammals 

 died out at the" end of the Oolite and did not exist dur- 

 ing the Cretaceous? We may be sure that the hiatus 

 occurs not in the records of the past, but in our frag- 

 mentary knowledge. 



I have said that intercalation does not prove descent 

 though it facilitates our progress towards that conclu- 

 sion ; but if between a fossil and a living form there 

 be intercalated a series of intermediate forms passing 

 gradually from the old type to the new, and those 

 gradations correspond with the age of the successive forma- 



