MAGNETIC OBSERVATIONS 19 



TORSION 



Two fibers were used until Thursday, October i, 1903, when they were found broken. 

 After this four fibers were used. New fibers were inserted March 27, 28, and 29, 1904, 

 the last serving through the remainder of the observations. Several attempts to use two 

 fibers alone failed, they being only sufficiently strong to suspend the torsion weight for but 

 part of a day. 



Observations for torsion were made before and after regular declination readings, ex- 

 cepting when the disturbances interfered. After the day's work the torsion weight was sub- 

 stituted for the magnet, the plane of detorsion was determined and the torsion weight was left 

 suspended until the next observations. When torsion had accumulated during observations 

 and made necessary a shift of torsion head, the effect of 90 of torsion is noted in the foot- 

 notes to tabulations. 



RECORDS 



The chronological program arranged by Dr. I,. A. Bauer could not be adhered to as 

 closely as might have been desired during the fall and winter, owing to prevailing strong 

 winds, when snow drifted to such an extent as to make travel to and from the observatory 

 both difficult and dangerous. Under these conditions two men holding on to a leading line 

 would struggle backwards to the observatory and dig out the entrance. This was filled again 

 in a few moments, imprisoning the observer until again liberated by outside aid. 



In order to economize space, the original notes have been tabulated as far as possible, with 

 the corresponding results. On pages 20 to 26 will be found the readings of the azimuth mark 

 under the headings Azimuth Mark No. 2 and Azimuth Matk No. i. In the column headed 

 Pointing will be found the letters B, A, which indicate respectively that the reading of the 

 azimuth mark has been taken before or after the declination observations of the day. Where 

 both letters appear the mean is published ; when none is given the azimuth marks were in- 

 visible on account of drifting snow or dense fog. 



The readings corresponding to the position of the telescope when pointed on the magnet 

 are given under the heading Circle reading of magnet. They are the mean of the readings 

 taken before, after, and sometimes during the period of declination observations. When the 

 telescope with the circle has been shifted during declination observation the values are omitted 

 in this table, and will be found on pages 26 to 31, with corresponding time. This last 

 mentioned table therefore shows when disturbances, so great as to require a shifting of the 

 horizontal circle, have occurred. The times of observation first following these shif tings of 

 horizontal circle in any day's work are also indicated in tabular summary of two-minute 

 declinations by asterisks (*). 



The circle reading of the true south is given for each day on which declination observations 

 have been made. Where there is no corresponding reading on an azimuth mark it has been 

 found by interpolating according to the dates and without considering the temperature effect. 



In the temperature columns will be found the maximum and the minimum thermometer 

 readings for the day, converted to Centigrade scale from the regular meteorologic record. 



The reductions of the individual two-minute readings for declination are tabulated on 

 pages 41 et seq. To make this tabulation quite clear the following specimen computation for 

 November i, 1903, is given : 



The scale readings for sh oom are (see page 71) left 59 d .8, right 65 d .3, 



the mean of which is 62". 5. 

 The axis for week ending November i, 1903 (see page 18), is 53 d .o, 



whence the difference, scale minus axis (S A), 9*. 5, which converted 



into arc (see constants, page 8) is + 14'. 9. 



