16 Land Magnetic Observations, 1905-10 



When, on account of clouds, there is prospect of difficulty in securing pointings 

 on the Sun, the order of observation is modified as may be necessary; thus, for 

 example, instead of making two pointings, vertical circle left, the instrument is 

 immediately reversed after the first pointing and another secured in'the opposite 

 quadrant. In cases where only one pointing was obtained, the observer determines 

 the horizontal and the vertical collimation of his instrument, by proper readings on 

 a mark; thus the necessary corrections may be applied. When the Sun is partly 

 overcast, or Avhen it must be sighted through the branches of trees, the desired 

 quadrant can not always be selected and the one available must be used, proper 

 record being made, of course, of the apparent limb observed; sometimes the same 

 limb must be taken on reversal. Occasionally in cloudy weather during the day, 

 the nights are clear, and sometimes in an emergency delay may be avoided by 

 observing a planet or a star, for azimuth and latitude. In general, however, the 

 observers are instructed to avoid observational work by night except in emergency 

 or when they have special facilities. 



Measurements of angles between prominent marks, reference points, etc., are 

 required where marks and reference points are available. In recording angular 

 measures on a series of marks the record indicates whether these readings precede 

 or follow the Sun work: the readings are invariably made forward and then back 

 with instrument reversed, in the reverse order, and so recorded, thus: IL, 2L, 3L, 

 etc., to nL, then nR, (n l)R, etc., to 2R, IR. 



Magnetic Work. 



Magnetometer observations. A complete set of magnetometer observations at 

 a station comprises (a) declination; (b) oscillations of magnet erect, not less than 10 

 differences of 50 or more oscillations, generally more, the initial arc of vibration 

 being, in general, not over 1.5, so as to render the correction on account of ampli- 

 tude of swing negligible; (c) deflections, both magnets erect, at two deflection 

 distances; (d) deflections, both magnets inverted, if construction permits, at two 

 deflecting distances; (e) oscillations of magnet inverted, not less than 10 differences 

 of 50 oscillations, generally more; (/) declination. 



Determination of the torsion coefficient is made at least once at each station. 

 Except for good reason, e. g., necessity of embracing, on cloudy days, an opportunity 

 afforded for astronomical observations, the corresponding observations of oscil- 

 lations and deflections are separated by the shortest possible interval. The same 

 thermometer is used throughout the intensity observations, being mounted in the 

 magnet-house during oscillations and in the deflection-box during deflections; in 

 deflection work the first temperature is noted at the end of the half set in order to 

 allow the thermometer time to take up temperature of bar; if temperature is chang- 

 ing rapidly it is read at the beginning as well as the end. Care is always taken to 

 stop up the thermometer hole in the magnet-house when the thermometer is re- 

 moved. 



A single declination value by magnetometer results from the mean of two scale- 

 readings (magnet erect), four (magnet inverted), and two (magnet erect), together 



