78 



BASE-LINE VALUES. 



CHAPTER II. 

 BASE-LINE VALUES, ANNUAL INEQUALITY, AND SECULAR CHANGE. 



10. Table II gives particulars of the values given by the absolute observations for the base line of the 

 Declination curves and the mean values for individual months as calculated and used. In the calculations 

 regard was sometimes paid only to the observations of the particular month, as in May, 1902 and 1903; 

 but in most cases the observations in adjacent months were also considered. Thus the calculated value for 

 June, 1902, is the arithmetic mean of the observed results on May 26 and June 30. In no case was 

 anything taken into account except the results of the absolute observations. 



The base line was assumed to have a constant value for each month. In passing from one month to the 

 next the change in the assumed value of the base line introduces a discontinuity, the amount of which is 

 shown in the last column of Table II. If the base-line value accepted for the second of two consecutive 

 months is higher than that accepted for the preceding month, then the first midnight of the second month 

 appears with a correspondingly higher Declination than the last midnight of the preceding month. The 



TABLE II. Values of the Declination Base Line. 



* Applied also during March and April, 1902. 



two hours are of course the same, and the apparent difference between the Declinations assigned to them 

 is wholly fictitious. 



The existence of such apparent discontinuities is of course undesirable. What they really imply in the 

 present case it is impossible to say. They are contributed to, no doubt, by observational errors ; but even 

 in the Antarctic it seems unlikely that errors of more than 5' or 6' would arise in absolute observations 



