166 THE PROBLEMS OF EVOLUTION 



recognizing that they are not sharply limited in 

 time if our concepts of evolution are sound. Every 

 character must have a beginning, but it must have 

 its beginning in preexisting structures. Novelty in 

 organisms can only be a modification of familiar 

 things. 



We trace the one-toed foot of the horse back to 

 a four-toed condition in ancestral species. It still 

 carries the vestigial metacarpals of two other digits 

 and the series of fossil remains connects these 

 vestiges with functional toes of various degrees of 

 development in a gradually receding series of ex- 

 tinct forms. And in the most remote known an- 

 cestor the fifth toe is present in a vestigial state. 

 Given the pentadactyl appendage as a primitive 

 character of the terrestrial vertebrates, we find its 

 predecessors in the paired fins of the Crossop- 

 terygii. The fins seem to establish an abrupt de- 

 parture, but there is rather good evidence for their 

 association with the fin folds of the Cephalochor- 

 data, and when we have arrived at this simple 

 beginning we are very near to the unmodified body 

 wall. Skeletal structures present a like sequence 

 in the vertebrates; bone, bone and cartilage, and 

 cartilage alone, establish a transition in the phy- 

 logeny of the skull, and all of these stages depend 

 ultimately upon the plastic mesenchyme. Such 

 sequences are also established in ontogeny, and 

 without regard to an individual's attitude to the 



