ORIGIN OF FLIGHT IN BIRDS. 



467 



that in most Archosauria the ulnar phalanges are reduced, 

 while the radial phalanges are retained or sometimes even 

 strengthened. He takes this to be a sign of the arboreal adapta- 

 tion of Proavis. The reduction cannot be denied, but again it 

 does not seem to be a sign of any arboreal specialisation. In 

 the one group of extinct reptiles that certainly was derived 

 from arboreal types this reduction does not take place, for in 

 Pterosaurians the last finger is very strong, and this reduction 

 does not occur in any of the arboreal mammals (5). It is well 

 marked only in Orocodilia, Dinosaurs, and birds. The primitive 

 Dinosaurs are, as is well known, bipedal (15). In the Orocodilia 

 the reduction of the shoulder-girdie and the curious elongation 

 of the carpal boi:ies indicate that formerly also in this group 

 the disproportion between the fore limbs and the hind limbs 

 was also decidedly greater (10), Considering that all those 

 mammals that are partly bipedal use the anterior limb for 



Text-figure 3. 



Footprints of digitigrade Triassic reptiles. 



(A) Sitpa^amopMS (somewhat modified) ; (B) Polemarchus ; (C) Ammoptcs; 

 (D) Platj/pterna. 



grasping, and that Dinosaurs and Orocodilia are terrestrial 

 forms with a strong posterior body, it should be investigated 

 whether the ulnar reduction is not due to the use of the manus as 

 a grasping organ during terrestrial and not during arboreal life. 

 The first change that is attained when both the as yet unmodified 

 anterior limbs are used together as a primitive grasping organ, is 

 the inward rotation of both palmar surfaces ; the next change 

 that must be expected is a strengthening of the radial side, for 

 in two palms that are half-way rotated the radial boi'ders of the 

 hands come always nearer together than the ulnar borders ; 

 consequently they are always exposed to much greater strain 

 than the ulnar sides of the hands. This greater strain will be 

 especially felt when the food is held with both anterior feet and 

 pieces are detached with the teeth. Such a manus must of course 

 difl^er strongly from the manus of every arboreal type, for in the 



