28 Land Magnetic Observations, 1914-20 



The corrections obtained by use of the graph are applicable immediately to the 

 azimuth or to the hour-angle of the observed body, but by proper consideration of the 

 subsequent steps in the reductioons, they may as well be applied at once to the ultimate 

 objects of the computation, viz, to the azimuth of the reference mark or to the correction 

 of the chronometer on local mean time. The rules given of the bottom of Figure 1 have 

 been made to meet this requirement. 



Magnetic Work. 



The determination of horizontal intensity at a field station with the magnetometer 

 involves measuring the period of one oscillation of the magnet by direct observation. 

 When the chronometer used has other than a zero rate, the observed period of one os- 

 cillation must be corrected for error caused by the rate of the chronometer. An observer 

 traveling in unmapped or poorly mapped regions, pending the discussion of his time ob- 

 servations upon the conclusion of the expedition, has generally only an approximation 

 to the true rate, and he is, therefore, obliged to make his computation using a zero or an 

 approximate rate; the correction for rate as finally adopted has then to be applied when 

 the office revisions are made. 



It is usually expedient to defer the decision as to the rate until the completion of the 

 expedition, when the observer has made his final report and has submitted all information 

 relating to geographical position such as maps and results of local surveys obtained in the 

 field. The adjustment from this material is often a troublesome process involving con- 

 siderable time and requiring careful analysis. In order that the revision of the observer's 

 results may not be delayed because of this work, it has been found practicable to apply 

 the necessary corrections on account of the finally adopted chronometer-rate to the 

 values of the horizontal intensity, H, and of the magnetic moment, m, as computed upon 

 the basis of the approximate or zero rate originally assumed. 



From the oscillation observations we have 



-^ (3) 



where K is the moment of inertia of the oscillating magDet and its suspension and T is the 

 period in seconds of one oscillation with all corrections applied, including that arising from 

 the chronometer rate. The relation between the observed period, T , corrected except for 

 rate to the true period, T, is 



T=tX- (4) 



s 



where s is the length of a true second of time and s' the length of a second as given by the 

 chronometer. If r is the daily rate of the chronometer in seconds, being positive for a 

 losing and negative for a gaining rate, then 



_s^ _ 86400 , 5) 



s 86400- r 

 whence, from (4), 



AT=T _ T JTr_ , . 



86400 

 From (3) we have, the subscript o indicating a value derived from oscillations only, 



AH =-^-AT (7) 



Am =-^AT (8) 



