32 



from the centre to the extremities, and, if the orig- 

 inal curvatures could be discharged, and the ex-- 

 tremities moved across each other, we should have 

 the first S replaced by another of reversed flex- 

 ures, and we should have the discharge of the orig- 

 inal curve of the rod so constrained by the nas- 

 cent one that its direction might be made entirely 

 perpendicular in both C C ; in a similar manner but 

 much more accurately than the discharge of the 

 simple C curve mentioned in § 19, was constrained 

 by the tube held over it. 



The power of the nascent new shape will ba 

 largely exhausted in thus constraining the direc- 

 tion of the spring from the old, but a portion 

 would remain to join the alternating points of ap- 

 plication in impressing the similarly shaped new 

 curve upon the rod. 



§ 30. If we examine more closely the lines of di- 

 rect and of reacting traction we think that the 

 manner in which a change of curvatures would 

 take place, if appropriate free articulations in the 

 course of the rod allowed the discharge of the first 

 cui've, may be explained as follows : 



The contrary curvature arising in each C is 

 constraining the direction of the old curvature 

 throughout, and incipiently altering it at the cross 



