MAGNETIC OBSERVATIONS 275 



REDUCTIONS FROM DECLINATION OBSERVATIONS AT TEPUTZ BAY 



Diurnal Variation 



For the purpose of determination of empirical formulae to express the diurnal variation in 

 the magnetic declination it was decided, in accordance with the general present tendency in 

 magnetic reductions, to include all of the observed values without elimination of any as 

 "disturbances". The few observations made between September 28 and October 4, 1903, are 

 not, however, included in the discussion as it was deemed that some time was necessary before 

 the routine and stability of observation could be properly established. 



The scheme of work carried out, as per program on page 17, was such that continuous 

 observations were made throughout one day of each week. On each of four of the remaining 

 days of the week observations were made for four hours continuously, and on one other day of 

 the week for eight hours continuously. These latter observations were so made as to cover, 

 when taken together, twenty-four hours numbered consecutively. There was thus obtained in 

 each week the equivalent of two days' continuous observation. The means of the thirty 

 observed values of the magnetic declination from one hour to the succeeding hour, as per the 

 tabulation of pages 41 to 274, have been taken as corresponding to the half hours local mean 

 time. Strictly speaking account should be taken of the chronometer corrections on local mean 

 time but as these were usually very small and varied in sign the resulting error is much below 

 the order of accuracy of the results arrived at and may, therefore, be disregarded. In order to 

 have the mean values correspond strictly to the mean epoch of the period under discussion, the 

 series was divided into intervals of four weeks each. Thus we have for each period, with few 

 exceptions, eight mean values for every hour, each resulting from thirty observations. The 

 means of these means have been taken as the hourly values applying to the mean epoch of the 

 period in qufestiou. 



The resulting hourly values of the declination for each interval and for the mean of the 

 whole period during October 4, 1903, to July i, 1904, at the Teplitz Bay station are exhibited 

 in the following tabulations, which are arranged according to local mean time, civil reckoning, 

 from midnight through twenty- four hours. Figures 5 to 15 show these values graphically, the 

 mean observed declinations being indicated by circles joined by broken lines ; the smooth 

 curves shown on these figures represent the computed values resulting from the analytical 

 expressions for the diurnal variations deduced from the same (see pages 290 to 291). 



