DARWIN AND PALEONTOLOGY 221 



been offered for such changes of proportion is 

 that of the selection of hereditary quantitative 

 variations.^ 



I am therefore myself inclined to regard long- 

 headedness or short-headedness in the vertebrates 

 generally as well as the similar phenomena of 

 long-footedness (dolichopody) or short-footed- 

 ness (brachypody) as in many cases caused by 

 the selection of changes of proportion; yet I 

 freely admit that we can not prove or demonstrate 

 this Darwinian hypothesis through paleontology. 



One direct paleontological reason may, how- 

 ever, be adduced in favor of the hypothesis of 

 selection of variations of proportion, namely, 

 changes of proportion do not fall under what I 

 call the law of ancestral control of variation. 

 Head proportions or foot proportions, or, in 

 fact, any other change of proportion can not be 

 regarded as controlled by ancestral affinity, be- 

 cause descendants of the same ancestors soon give 

 rise to very different results. For example, a 

 primitive intermediate (or mesaticephahc) form 

 of skull does not at all control the form of skull 

 which may be derived from it ; the animal is free, 

 as it were, to evolve into one of many different 

 kinds of head forms. The point is that hered- 

 itary predetermination does not appear in the 

 evolution of proportion and of form as I shall 

 show that it does appear in the evolution of cer- 



'■ The fact that this throws little light on the origin of dolicho- 

 cephaly or brachycephaly in the human species appears to throw 

 the selection hypothesis again into doubt. 



