490 



SCIENCE 



[N. S. Vol. XL VI. No. 1194 



second set is determined by starting a stop- 

 watcli as the first reading of the first set taken, 

 and stopping it at the first observation of the 

 second set. This time divided by the known 

 period of the pendulum will fix the number of 

 oscillations from the first of one set to the 

 first of the other, i. e., it will give the os- 

 cillation number of the first elongation of the 

 second set, the initial elongation of the first 

 set being taken as zero. Thus knowing the 

 distance of the scale D from the center of os- 

 cillation 0, and the intersections of the two 

 vertical planes at D, we get the angle be- 

 tween them ; and from this and the time inter- 

 val the angular velocity of the rotation of the 

 plane of vibration follows at once. 



The first attempts to use this method were 

 made with the slit about two meters from the 

 center at C, and the scale sis meters away. 

 The observations were made on the first dif- 

 fraction minimum to one side of the pattern, 

 but the decay of the amplitude of vibration 

 introduced here an undeterminable correction 

 which was too large to be neglected. The 

 final procedure was to put the arc about six 

 meters from the center and to bring the scale 

 to two meters. Eeadings were then made of 

 the edge of the central dark shadow — the 

 bright line in the middle of the shadow being 

 too faint for quick reading. Under these cir- 

 cumstances the variation of the width of the 

 central shadow, even in its extreme positions, 

 was negligible. 



The Apparatus. — A turned leaden sphere of 

 mass 4.8 kilograms was suspended from a 

 roof joist of the laboratory by a long steel 

 piano wire 0.39 nun. diam. Attached to the 

 wire, so that its shadow would cross the scale 

 D at each oscillation, was a small ball of wax. 

 As the screen was about a meter above the 

 floor and the arc about 20 centimeters, this 

 shadow was at its highest point at one max- 

 imum elongation of the bob and at its lowest 

 at the other. By noting the motion of the 

 shadow of the wax ball at the ends of its path 

 one could detect any tendency to elliptical 

 motion of the bob. The prevention of such 

 motion is, of course, one of the difficulties in 

 securing good results. 



The period of the pendulum was 7.50 sees. 



To start the oscillation the bob was drawn 

 back 40 or 50 cm. from its equilibriiim posi- 

 tion and held there by a belt of thread that 

 passed about its equator and through a small 

 horizontal pulley, which latter was fastened 

 to a standard by the thread which was to be 

 burned in releasing the pendulum (see Fig, 

 3). The object of the pulley was to prevent 

 torsional strain in the wire, but as the re- 

 storing couple was so small for the wire in 

 question it was found best to place a mark on 

 the sphere after it had been hanging at rest 

 for some time, and to adjust the ball in its 

 belt so that the mark was at its original 

 azimuth. Next, to damp out side motion of 

 the bob the following device proved efiicient; 

 a flat disc of cork (about 2 cm. diam.) was 

 fixed centrally on the inside of a light tin 

 dish (top of a coffee can, 11 cm. diam. See 

 Fig. 3) and this was floated on cylinder oil in 



a larger vessel that was carried on a table that 

 could be racked up and down (the front of a 

 projection lantern). This system was placed 

 centrally under the bob in its deflected condi- 

 tion, and was raised until the cork just 

 touched the sphere. The slight friction be- 

 tween them caused the dish to move with the 

 bob, so that the oil quickly damped the resid- 



