232 Kansas Academy of Science. 



side by a small cord attached to a staple fixed as near as could be 

 estimated to its center of oscillation. The thread is fastened to the 

 wall and the bob, thus displaced, is left until it comes entirely to 

 rest. Then the string is burned off and without any lateral push 

 the pendulum falls, simply under the impulse of gravity. The 

 smoked-glass plate rests on a platform, which lias a mechanism for 

 raising and lowering with parallel motion, and permits the tracing 

 needle to follow the plate two or three feet or any other desired 

 distance. 



There are several disturbing agencies which prevent the pendu- 

 lum from tracing a straight line as it swings, and make it move in 

 an ellipse of greater or lesser eccentricity. The most troublesome 

 of these is the action of air currents, which cannot be wholly 

 avoided. Even if the room were tightly closed, the variations of 

 temperature and movements of the observer would cause fluctua- 

 tions of air which would sensibly affect the pendulum. 



Another difficulty is the elasticity of the suspending wire. This 

 is bought in coils and it is difiicult so to straigthen it that it will 

 not tend to fall in spiral twists which are not equally elastic in 

 every direction. Then there is lack of perfect rigidity in the point 

 of suspension, which must also fail to represent a true mathemat- 

 ical point. There is no better plan for suspension yet devised 

 than the method described above, of allowing the wire to pass 

 tightly through a hole in a metal plate. 



The author once thought it important to provide a plan for ro- 

 tation of the wire, but became convinced that this refinement was 

 entirely useless. Aside from these mechanical imperfections there 

 is another and an insurmountable object, which prevents the pen- 

 dulum from swinging backward and forward in the same path. 

 When the pendulum is drawn aside at its starting it must partake 

 of the rotary motion of the earth, and this, combined with the 

 gravitational force, is sure to give it an elliptical course. In se- 

 curing the graphical representation of these vibrations, there is still 

 another disturbing element, for our needle-point can seldom be 

 brought into the true axis of the pendulum. Where this needle- 

 point is not in the true axis its friction against the smoked glass 

 tends to give the bob a rotary oscillation, and, in the experiments 

 hitherto made, this rotary oscillation is quite in evidence. It gives 

 the outside edge of our blue-print a curious wavy border, and the 

 time of these oscillations is many fold greater than the regular, 

 swings of the pendulum. The impelling force is here the torsional 

 elasticity of the wire, which, acting on the heavy weight, gives a 



