228 'Experimental Transformation of Species 



quite a different problem to find out whether the successors 

 were also descendants, notwithstanding the marked differ- 

 ences between one type and the next. 



It must be recognized that from the nature of the case 

 it is quite impossible to find direct evidence. Even if all the 

 forms successively inhabiting an area were of exactly the 

 same type, the descent of the later ones from the earlier ones 

 would be highly probable but only a matter of inference. 

 We could have a given house occupied for a hundred years 



Fig. 57. The Evolution of Horns in Deer 



Fossils of the earliest representatives of the deer family show simple horns. In 

 later and later fossils, the horns branch more and more. This fact of the pro- 

 gressively more complex horns corresponding to the later forms is parallel to the 

 increasing complexity of the horns with advancing age of the individual deer 

 today. After Romanes. 



by Robinsons, but by a succession of different families claim- 

 ing and recognizing no relationship to one another. On the 

 other hand, the residents in a house might offer the post 

 office a succession of distinct family names and yet be always 

 direct descendants of the original occupants. To find 

 whether succeeding forms are modified descendants it is not 

 sufficient to arrange more or less continuous series. There 

 can conceivably be intergrading or overlapping of unrelated 

 forms. It is possible, in fact, to find well graded series of 

 characters that are known to be of independent origin. 



