458 K D. Preston— Study of the Earth's 



noted with a chronometer beating whole seconds. This re- 

 quired the invention of an entirely new kind of apparatus. It 

 was evident some means must be had by which the coincidence 

 could be noted optically. For this purpose the chronometer 

 was made to open every second the armature of a relay to 

 which was attached an upright thin piece of metal. This 

 metal was perforated by a thin slit which by the movement of 

 the armature passed before a iixed slit a short distance in front 

 of it, so that a light suitably disposed gave a flash every time 

 the chronometer broke the electric current. The apparatus 

 was placed about ten feet from the pendulum, and was so 

 adjusted that the beam of light from the slit fell on two 

 mirrors, one of which was on the pendulum and one near by 

 it. The former was therefore movable by the motion of the 

 pendulum while the latter was stationary, From these mirrors 

 the beam of light was reflected back into the observer's teles- 

 cope. When the pendulum was at rest, the observer saw two 

 illuminated slits every second in the held of view, but with 

 the pendulum in motion the flash that came from the pendu- 

 lum mirror could only be seen when the pendulum happened 

 to be near its equilibrium point. It is evident that if the pen- 

 dulum makes exactly two oscillations for every second of the 

 chronometer, the relation of the images will not change, and 

 we shall see a double flash every second in the middle of the 

 held of view. But if the time of oscillation of the pendulum 

 is slightly different from a half second, it will not return quite 

 to its former position by the time the next flash occurs, and 

 we shall have its image displaced with reference to the one 

 from the fixed mirror. Here then we have two necessary con- 

 ditions for the appearance of the flash from the pendulum. 

 First the chronometer must open the slit, and second, the pen- 

 dulum at this instant, which has a duration of about 1 /100th of 

 a second, must be in such a position that the image of the slit 

 will be reflected back into the observer's telescope. We only 

 need now a point of reference to which the motion of the mov- 

 able image may be referred. This is furnished by the flash 

 reflected back from the stationary mirror, and as the image 

 from the pendulum is seen to pass slowly across the field of 

 view, the time is noted when the two images coincide. For 

 pendulum A, the time of oscillation exceeded half a second by 

 0*0066 seconds, which gave a coincidence interval of 6 m 15 s . 

 This was adopted for the other pendulums. Inasmuch as it 

 was contemplated to use these instruments in all parts of the 

 United States and at all altitudes, it was necessary to consider 

 the effect of a change of latitude and elevation on the coinci- 

 dence interval, because a period might be chosen at Wash- 

 ington, which would make the intervals too long for conven- 



