165 



the actual change of the spinal convexities (§ 61), 

 the movement of the diagonal feet of appui, in 

 the backward trot, must be more nearly synchro- 

 nous than in that of progression. (4.) The action 

 of the lower jaw must, in proper backing, be that 

 of yielding, instead of forcing, at the moment of 

 the spring, the alternating joint leading the work- 

 ing one. 



(5.) In hacJdng (§ 70) the feet leave the ground by 

 the outer bearings and take it again by tJw inner 

 bearing, exactly the reverse of progi-ession. 



§149. The Halt. In the "the third result" of 

 forcing the winding lines in the snake (§ 71), we 

 endeavoured to describe that of bringing on 

 the formation of one line by induction from the 

 continued forcing of the other, after the alloioed 

 counteractions had been already formed ; also to 

 show that so doing would bring on an equaHzation 

 of gathering in the two sides, resulting finally in the 

 ^' superimposition of twists," a formation in which 

 all four winding lines are developed at the same 

 time, and consequently all lateral curvatures sup- 

 pressed. 



As the actions of halting are identical with those 

 on which our theory of " setting up " is founded, 



8 



