General Methods of Work 21 



1909, and not simply Nov. 15, 1909, in this way providing a check against blunders 

 in recording dates. Information concerning conditions under which observations 

 are made, both as regards the observer and his surroundings, are always found 

 useful by the office computer and are given under "Remarks" on the respective 

 sheets. Specimen records of observations are given in a subsequent section. 



Computations. Before leaving a station, the observer carries the computations 

 far enough to make sure that there is nothing radically wrong with the observations. 

 Thus, in general, for good work two consecutive sets of azimuth should agree within 

 1 minute of arc, and morning and afternoon sets within 2 minutes of arc; greater 

 differences are usually due to lack of adjustment or level of the theodolite, to a 

 blunder in pointing on wrong limb of the Sun, or to erroneous value of latitude. 

 In case the diflference between morning and afternoon azimuth amounts to more 

 than 5 minutes, the observations are repeated. The two sets of declination should 

 not differ more than 2 or 3 minutes when approximate allowance is made for diurnal 

 variation. The average mean time of 70, 100, or whatever number of oscillations 

 is used, should in general not differ more than 0.5 second in the two sets and the 

 deflection angles should in general agree in the two sets within 0.5 minute of 

 arc for the short distance and 0.25 minute of arc for the long distance; these limits, 

 however, depend on the instrument used. The attempt is to secure an accuracy 

 of at least one-thousandth part of the value of the magnetic horizontal intensity. 

 If the dip results for the individual needles, when reduced to standard, differ much 

 over 2 minutes from each other, the observations are repeated. 



The observers soon acquire the practice of carrying out observational means as 

 the observations progress, thus making possible the immediate detection and 

 correction of blunders. When for some good reason the computations can not be 

 made before the observer must go to his next station, a close inspection of his 

 observational means generally suffices to indicate whether or not the observations 

 are satisfactory; for example, a close scrutiny of the means as called for on the 

 azimuth forms will point out erroneous work. 



In general, five-place logarithms are used in the reductions. For the azimuth 

 computations, the means of circle readings are carried to tenths of minute of arc, 

 the times to tenths of seconds; in the time computations, the hour angle in arc is 

 taken out to whole seconds. In the declination reductions, the mean scale readings 

 are carried to hundredths of a scale division, and the balance of the computation 

 to tenths of minute of arc. For the intensity work by the magnetometer, the mean 

 time of one oscillation is computed generally to four decimal places, the mean 

 temperature to hundredths of a degree, and deflection angles to the nearest second 

 of arc. The dip computations are carried to tenths of a minute of arc. The local 

 mean times are computed to the nearest minute. 



Reports. Each observer, in addition to making a brief monthly report of his 

 operations, upon the completion of a field campaign, makes a report on the work 

 executed, covering all matters of interest and containing information and sug- 

 gestions for guidance in future expeditions. 



