80 



MONTHLY MEANS. 



12. Tiible FV shows two sets of mean values of Declination for individual months derived from the 

 hourly readings, accepting the base-line values given in Table II. The first set of data are derived from 

 all the days of complete registration ; the second set are derived from a smaller number of days, selected 

 as the least disturbed of the month. The differences between these two sets of values given in the last 

 column would be unaffected by any alteration in the monthly values accepted for the base lines. The 

 smallness of these differences seems to justify the conclusion that they are not seriously affected by the in- 

 cidence of magnetic disturbances or by uncertainties in the measurements of individual days' curves. Whilst 

 the differences are small, the all-days' mean appears in excess in so large a majority of cases as strongly to 

 suggest that the phenomenon is not a purely accidental one. Differences between mean values derived 

 from all and from quiet days have been observed elsewhere, though not of so large a size. 



TABLE IV. Mean Monthly Values of Declination (from the Curves), employing the Base-Line 



Values accepted in Table II. 



13. Table V gives the mean monthly values of the Horizontal Force as derived from the hourly values 

 in days of complete registration, accepting the base-line values given in Table III. 



TABLE V. Mean Monthly Values of Horizontal Force (from Curves), employing the Accepted 



Base-Line Values in Table III. 



