CiiAr. 7] GRAVITATIONAL METHODS 113 



inclosed in an evacuated glass bulb not much larger than a radio tube 

 (see Fig. 2-1), consists of a fused quartz rod 4 mm in diameter and 60 mm 

 in length, extends into a pin, P, for observation or photoelectric recording, 

 and is fastened at the bottom to an elinvar spring, E, which at the thinnest 

 point has a thickness of only 0.02 mm. An ingenious arresting mechanism 

 clamps the pendulum by a slight movement of the diaphragm, D. In this 

 manner, the vacuum inside the tube is not disturbed. The period of this 

 pendulum is about 6 to 7 seconds; the time required for a single observa- 

 tion is about 4 minutes. A 40- to 60-minute observation period gives 

 better than one milligal accuracy in gravity. 



C. Time-Determination and Time-Signal-Transmission Methods 



For an accurate determination of the pendulum period some sort of a 

 standard timepiece must be used, such as a contact chronometer, a contact 

 clock (Riefler), or a gravity pendulum. None of these (with the possible 

 exception of a well-protected gravity pendulum at a central station) 

 retain a sufficiently constant rate and must be compared with absolute 

 time standards. This comparison may be made (1) astronomically, with a 

 zenith telescope, (2) by recording of observatory time signals transmitted 

 by wire, or (3) by radio. The following discussion of time-determination 

 and time-signal-transmission methods will include a description of pro- 

 cedures used for transmitting pendulum oscillations from a base to a field 

 station or vice versa. 



1. Astronomic time determination is now used in emergency cases only 

 when reception of time signals is impossible. With a zenith telescope the 

 time is determined when a star (or the sun) passes the astronomic meridian. 

 At that instant the hour angle of the star is zero and its right ascension is 

 equal to the local sidereal time; therefore, the "time correction" of the 

 chronometer is right ascension minus chronometer time. 



2. Reception of observatory time signals. In most of the U. S. Coast and 

 Geodetic Survey pendulum work until about 1932 the telegraphic noontime 

 signals of the U.S. Naval Observatory were used. They were recorded on a 

 chronograph, together with the beats of the contact chronometer. Relays 

 were employed throughout to save the contact points in the chronometers, 

 since their time lag does not affect the chronometer rates as long as it 

 remains the same in successive time signal observations. If the telegraph 

 office is too far away from the pendulum room where the chronometers are 

 located, a hack chronometer is compared with the stationary chronometers, 

 then carried to the telegraph office, and afterwards compared with the 

 stationary chronometers. 



It is now the more common practice to record radio time signals on a 

 chronograph toegether with the beats of the comparison chronometer. 



