78 THE LESSON OF EVOLUTION 



both to be reptiles. But the theriodonts form such 

 .a close connection between reptiles and mammals 

 that it is difficult to decide which of the two to call 

 them ; while they are also, in the lower forms, related 

 to the Amphibians. Between the Amphibians and 

 the fishes there is a considerable gap, as yet un- 

 bridged ; and palaeontologists are undecided as to 

 whether the former are descended from the Dipnoi 

 or from the Crossopterygii ; for the origin of the five- 

 toed limb in the amphibians is obscure. Also we feel 

 no certainty about the course of development among 

 the fishes themselves, and about the line of descent 

 of the lowest fishes from invertebrates. 



The reason for this is obvious. From the amphi- 

 bians upwards we have animals with a hard 

 skeleton, easily preserved; while from the 

 amphibians downward the skeleton is cartilaginous. 

 Our clue is lost, and our knowledge is almost entirely 

 confined to those exceptional animals which de- 

 veloped a hard dermal armour. The whole of the 

 facts may be summed up in this sentence : The more 

 complete our knowledge of extinct animals, the 

 clearer is the evidence for development. This really 

 amounts to a proof of the theory, which was origi- 

 nally arrived at by biologists from a study of living 

 animals, and is now confirmed by palaeontologists 

 through a study of the animals which formerly 

 inhabited the earth. 



SPECULATIONS 



Migrations from Land to the Sea, Perhaps the 



