C. B. WARRING. 159 



Again : Remove the wire, d e, support the end of the 

 frame, d, by a cord from the ceiling, and set the wheel 

 in rapid motion. 



The instrument will begin slowly to revolve around 

 a b, but will fall before making a revolution, probably 

 before going 90°, the distance depending upon the 

 length of the string and its position. 



This experiment proves that the cause of the swaying 

 is in some way connected with the rod d e. 



The rationale, I think, is dependent upon that law of 

 which we have had so many illustrations, viz., accel- 

 erating the horizontal movement causes the instrument 

 to rise. If I push it towards me, g gets a horizontal 

 momentum, which, when at the bottom, tends to throw 

 that part towards the observer, i. e., to cause it to re- 

 volve around an axis passing through d a, and hence 

 to throw h away from him. 



If, therefore, h was leaning forward, the effect of my 

 pushing would be to throw it, h, back to the vertical, or 

 even beyond. If it went beyond, then the tilting would 

 be the other way, and the push should be applied so as 

 to cause d to recede from the observer, and this would 

 send h back to the vertical. 



The question remaining is, Where does the accelera- 

 tion come from ? 



Of course, it is impossible to place the instrument in 

 stable equilibrium. The moment d e ceases to be ver- 

 tical, d falls with a velocity depending upon the short- 

 ness of the wire, while h would fall at a slower rate, be- 

 cause of its greater distance from the supporting points, 

 on the well-known principle of the pendulum, the 

 longer the rod the slower the motion (inversely as the 

 square root of the length). Therefore, as d inclines 

 toward the observer, the rod d e pushes d more rapidly 

 than it would go under the influence of h alone ; in other 

 words, it accelerates d, and consequently the instru- 



143 



