222 DRATTGHT AND HARNESS. 



curving backwards the upper part of the collar, which 

 hides the defect partly, and if not carried to the excess 

 of impeding the play of the top of the shoulder-blade, 

 is harmless. The other offers a curious instance of 

 misplaced ingenuity, and is of rather recent date — we 

 mean, the strap led back from near the top of the collar 

 to the trace-buckle, and which converts the front end 

 of the trace practically into a fork, whose points are 

 attached to the hames opposite to the two movable 

 ends of the shoulder-blade, so that in fact the play of 

 this bone is effectually checked at both ends alternately. 



The fair and inevitable conclusion to be drawn from 

 all this is, of course, that the trace should be attached 

 as nearly as possible opposite to the immovable part 

 of the shoulder-blade — that is to say, to its centre. A 

 very intelligent London harness manufacturer showed 

 the author very recently a number of collars and hames 

 constructed precisely on this principle, and assured 

 him that all the better houses in the trade had adopted 

 it for some time, so that all we can claim here is the 

 credit of showing the wdiy and the wherefore of what 

 ought to be, but unfortunately is not, generally adopted. 



The shoulder-blade we may now neglect for a mo- 

 ment, and go on to consider what other members are 

 brought into action, and in what direction the effort is 

 made by a horse leaning up against the collar in 

 draught. Now^ it is quite evident that in this case 

 the fore legs are much more completely simple bearers, 

 and the hind ones pure propellers, than under the 

 saddle ; and turning to fig. 4, p. 41, which was con- 

 structed altogether without reference to the question 

 before us, we find that the line R S \^ that in which 

 the propulsion is effected by the hind leg ; and now 

 turning again to fig. 1, and imagining a straight line _ 



