50 SEATS AND SADDLES. 



is tilted up, and at every movement of the horse and 

 rider the whole saddle is shoved forward till stopped} 

 hj the withers, which it will probably wound ; and 

 then it is either the groom's, or the saddler's, or the 

 horse's fault, and the saddle is thrown aside and some 

 nevr patent contrivance adopted, which of course does- 

 not remedy a defect that depends on the rider himself. 

 We may now go a step further. Suppose the sadr 

 die be placed with its centre exactly over the com-l 

 bined centres of gravity and motion (line B F, fig. 4),\ 

 and the rider in the centre of the saddle, there will be, \ 

 first, an equable distribution of the combined weight 1 

 of horse and rider on all four legs, both in a state of 

 rest and action ; secondly, the movements of the horse, | 

 centring in this point, have the least possible ten- | 

 dency to disturb the seat of the rider or the position j 

 of the saddle ; thirdly, the weight of the rider being 

 equably distributed over the whole surface of the sad- 

 dle in contact with the horse's back, is therefore less 

 likely to injure any one portion of this ; nor does it 

 convert the saddle into a lever, and shove it forwards ■ 

 or backwards. Again, let us suppose the saddle as ' 

 before, but the rider sitting altogether at its hinder 

 end for instance, and there will be, first, the horse's 

 equilibrium destroyed ; secondly, the rider himself, 

 being nearer to the hind legs, will first receive an im- 

 pulse from the direction R S, and be thrown forward 

 till he meets that coming from the direction P Q, and 

 these two forces, instead of resolving each other from 

 one common point into -their sum total, neutralise j 

 each other partially in successive shocks at the expense 

 of the horse's legs. It will be said that the use of 

 the stiriTips is to prevent the rider being thus thrown 



