284 EXPLORATION GEOPHYSICS 



this apparatus are properly adjusted, dark and light bands will appear in the field of 

 the telescope, due to interference of the sodium light waves of the two beams. 



The slight motion imparted to the case by the swinging pendulum shifts the mirror 

 on the case, making a small variation in the length of the path of one of the beams. 

 This causes the interference bands to shift back and forth. The amount of the shift 

 may be measured by a scale in the field of the telescope. The movement of the edges 

 of a band divided by the width of the band (in scale divisions) gives the movement of 

 the receiver in units of band width. 



Tests with pendulums swung on supports of different degrees of flexibility will 

 show the relation between the movement of "fringe" bands and the resulting error in 

 the period of the pendulum. Experiments indicated that for the above-described equip- 

 ment a movement equal to the width of 1 band produced a change of 173 in T in units 

 of the 7th decimal place. This constant was determined with a pendulum swinging 

 through a 5 mm. arc on the scale. All flexure observations are reduced to this arc 

 before correcting T for flexure of the support. 



The correction to the period is D • 173, where D = the displacement in fringe 

 bands per 5 mm. of arc. It develops that an error of 0.014 fringe of sodium light in 

 observing flexure may cause an error of 0.001 in the value of g. 



Flexure of the support is found in some types of apparatus by observing the oscil- 

 lation acquired by a special pendulum in a separate place on the supporting framework. 

 When a run is started this flexure-measuring pendulum is at rest. It is set in motion 

 by the vibration of the working pendulum of the equipment. Flexure and the correction 

 for it can be determined in terms of the period of swing imparted to the special 

 pendulum. 



To insure a rigid support for the base chamber, a hole about 10 inches in diameter 

 and 12 inches deep is dug at each station. Plaster of Paris is placed in the hole and 

 wetted with water. When the proper consistency has been obtained, the base chamber 

 is set into the plaster and leveled with the small universal level bubble in the base. 

 Most plasters will set rigidly within about thirty minutes. At permanent stations, 

 concrete piers are recommended as a support for the apparatus. A thin layer of plaster 

 of Paris is used to bond the base to the pier. 



Example. — A certain set of pendulum observations made in 1901 on the island of 

 Sumatra showed the first coincidence recorded at 9h 59m 03s and the final one at 

 16h 54m 46s, covering a total time of 6h 55m 43s, or 24,943 seconds. The approximate 

 length of a coincidence interval was 181 seconds. 



The number of coincidence intervals was verified by dividing 24,943 seconds by 

 181, which gave 137.9. As the number of coincidences must be a whole number, 138 

 was indicated. The true length of the coincidence interval (24,943 S./138) was 180.75 

 seconds. 



The uncorrected period = 0.501386 sec, which shows the accuracy of the method. 



To this figure the following corrections were applied : 



Uncorrected Period 0.5013869 



Corrections 



Arc -5 



Temperature -436 



Pressure -}-9 



Rate (Chronometer) -f-128 



Flexure -6 



Corrected Period 0.5013559 seconds 



As previously described, the value of gravity can be found, for the above field 

 station, by comparing the period of the particular pendulum used with its period at 



