856 



THE AMERICAN NATURALIST [Vol. XL 



of little use in general selection and competition with other animals. 

 Thus in the case of horns, of tusks, and of canines, bv favoring 

 the males in which they are most strongly developed they cause 

 an incidence of selection on characters which are useful in sexual 

 selection only. 



Extreme specialization in sevcial nicnil)ers of the titanothere 

 family took the form of dominance of the liorns. As we have 

 seen, the horns first ap])eare(l alike in hotli sexes as rudiments or 

 small horns but gradually they hec;niH> male chai'acters. and were 

 undoubtedly of advantage ti) the males in their >exnal eombats 

 for the possession of the females. Thus a constant selection of 

 the individuals with the largest horns may have l.e.Mi in process. 

 This incidence or main emj)l)asis of natural -.election on characters 

 which were useless for feeding purposes may havt> been the cause 

 of the non-evolution of the teeth. 



The force of this generalization is, however, weakened by the 

 fact that in other Titanotheres, such as the genera Titanotherium 

 and Megacerops, the horns were relatively small, yet these animals 

 became extinct at the same time as the large-horned genera, 

 Brontotherium and Syndjorodon. 



