C. 15. WARRING. 199 



and thereby, cause the gyroscope to move towards the 

 observer. The result is that the plane of vibration is 

 constantly changing. To show the projection of the 

 arcs, I attached at m, fig. 48, a conical pail with a small 

 opening at the bottom, and partially filled it with fine 

 sand, thoroughly dried to make it flow more readily. 

 The sand coming out as the gyroscope swung back and 

 forth left a line on a sheet of pajjer placed as closely as 

 possible beneath it. This is approximately copied in 

 fig. 50. To see the effect, one must follow the lines in 

 the order of the letters. Starting at A, it swings around 

 in an arc of a circle, passing through the centre and on 

 to B. At B there is a slight fullness in the curve, which 

 then becomes convex the other way, and passes on to C. 

 Here is a similar change in the sign of the convexity, 

 and, again going through the centre, the line reaches D. 



When a string is used instead of a rod the curves as- 

 sume a greater variety of forms, but as no new condi- 

 tions are involved it seems needless to pursue the matter 

 further. 



By way of contrast, as well as to exhibit the lack of 

 " inherent power " on the part of a rotating body to re- 

 sist change of direction of the plane of rotation, I sus- 

 pended my pendulum (fig. 48) by two cords attached at 

 o, and fastened at their upper ends to two staples, some 

 two feet apart. I released the end A, and drew a cord 

 from one staple to the other, passing it securely around 

 the irpper end of the rod, so as to prevent its slipping 

 either to right or left, my gyroscope thus being at the 

 vertex of an isosceles triangle, while the rod represented 

 the perpendicular to the base. I then pulled the " bob " 

 one side and noted the beats. Then I set the wheel in 

 motion and again pulled it one side and let it go. I 

 could see no difference ; the rotation neither accelerated, 

 nor retarded, the movement of the pendulum. 



183 



