IMPERFECTION OF THE RECORD, 93 



rocks have no skeleton. This is the more important, 

 because in very many cases the animals which we 

 must look for as ancestral types, could have pos- 

 sessed no skeleton. For instance, the original 

 vertebrate ancestor could not have had a skeleton, 

 and we can therefore have no hopes of finding any 

 record of its existence among fossils. (2) Fossils 

 can only be deposited during the accumulation of 

 sediment, and, therefore, during periods of slow 

 subsidence at the localities in question. All periods 

 of elevation, or of rapid subsidence, will be almost 

 or entirely unrepresented in the geological record. 

 How great these gaps may be, we have no means 

 of determining, but they are doubtless enormous. 

 (3) Fresh-water and land animals can only be pre- 

 served by being carried to the sea, or some inland 

 lake which is filling up ; or by being deposited in 

 the river beds. Evidence of this is found in the 

 great scarcity of fresh -water fossils, and particularly 

 in the remarkable scarcity of insects in all rocks a 

 group which must have existed in numbers, but 

 whose light bodies prevent them from sinking to 

 the bottom of the water. (4) Many of the rocks, 

 particularly those of the earlier ages, have been 

 subjected to such great metamorphic changes, that 

 whatever fossils they might once have contained 

 have been entirely or partially destroyed. These 

 four factors, with other minor ones, have appeared 

 sufficient to geologists to justify the belief in almost 

 any amount of imperfection in the geological record. 

 This imperfection is certainly growing less with 

 the continual study of new localities, for many gaps 



