READING RIDDLES OF THE ROCKS 125 



may confirm or refute this hint ; for a blunt claw 

 could not be used in tearing prey limb from 

 Hmb, nor would a double-edged tooth, made 

 for rending flesh, serve for champing grass. 



But few boneS of the feet, and especially the 

 fore feet, are present, these smaller parts of the 

 skeleton having been washed away before the 

 ponderous frame was buried in the sand, and 

 the best that can be done is to follow the law 

 of probabilities and put three toes on the hind 

 foot and five on the fore, two of these last 

 without claws. The single blunt round claw 

 among our bones shows, as do the teeth, that 

 Triceratops was herbivorous ; it also pointed a 

 little downward, and this tells that in the hving 

 animal the sole of the foot was a thick, soft 

 pad, somewhat as it is in the elephant and rhi- 

 noceros, and that the toes were not entirely 

 free from one another. There are less than a 

 dozen vertebrge and still fewer ribs, besides 

 half a barrelful of pieces, from which to recon- 

 struct a backbone twenty feet long. That the 

 ribs are part from one side and part from an- 

 other matters no more than it did in the case 

 of the leg-bones ; but the backbone presents a 



