264 TOWARDS THE HUMAN FORM 



immanis, which was nearly sixty metres in length. Morocaurus 

 grandis, on the contrary, was half the size of Bronto- 

 saurus, and had only four sacral vertebrae instead of five. 

 Another of their contemporaries in Colorado, the celebrated 

 Dipiodocus longus, completely reconstructed by Professor 

 Holland, and a magnificent cast of which was given by Carnegie 

 to the Museum d'Histoire Naturelle in Paris, measured 26metres 

 in length. Its head, not so small as that of Brontosaurus , had 

 somewhat the aspect of a horse ; its lower jaws were provided, 

 in the front only, with long incisor-like teeth, each accompanied 

 inside by a row of replacing teeth. 



The form of the teeth in all these large Reptiles shows them 

 to have been herbivorous. If the connexion, previously pointed 

 out, between the multiplication of the vertebrae and the part 

 taken by the trunk in locomotion be recalled : if to this we add 

 that the caudal vertebrae of Dipiodocus was provided with 

 bones arranged chevron fashion, and each having two sym- 

 metrical horizontal rafter-like supports, indicating that this 

 tail must have been planted on the ground and used for 

 propulsion — then we are led to the conclusion that the Sauro- 

 podians lived in dense jungle, through which they had to thread 

 their way by separating the growth with movements of their 

 long necks and then pushing into and through it by the leverage 

 of their tails, their limbs permitting them to raise themselves off 

 the ground where the tangle of branches was thickest. The 

 shape and small size of their feet excludes the notion, sometimes 

 advanced, that they were marsh animals, and the position of the 

 nostrils, on and not at the extremity of the muzzle, a character 

 common to Sauropodians and aquatic animals, is even better 

 explained by the use they made of their heads in forcing the 

 branches apart. This action alone would mechanically push 

 the nostrils back, the more so because if they had been situated 

 at the end of the snout they would have been constantly torn 

 and blocked by thorns. 



All the large Reptiles above mentioned belong to the Upper 

 Jurassic, but analogous animals must have been living much 

 earlier, from the Lower Trias, in fact. In the sandstones of 

 Fozieres near Lodeve, which date back to that period, and in 

 those of Connecticut, imprints of pentadactyl feet with a 

 separated pollex actuallyhave been observed. These cannot have 

 been made by Stegocephala, in whom four digits is a constant 



