88 EEi-ORT— 1898. 



made cyclic, and from all days, b-c, the greatest and least monthly values 

 are given for each element, both as referred to months and years. As in 

 Table VIII., the mean of the differences between the greatest and least 

 values differs for the mean for months from that for the mean for years in 

 all elements, in regard to which the remarks made in speaking of Table 

 VIII. apply also here. 



The mean values of b — c are most marked in declination in winter, and 

 in horizontal and vertical force towards or in summer. The deviations 

 between values of diurnal range, however determined, whether from quiet 

 days as observed, or made cyclic, or from all days, are in a sense small as 

 compared with the annual march of each element from its winter value to 

 its greatly increased summer value, and back again to winter value. In 

 the grouping for years the values oi b — c for vertical force show a progres- 

 sion from year to year for which there seems to be no sufficient explana- 

 tion. The mean diurnal ranges of this table are throughout combinations 

 of the diurnal ranges for individual months, formed from the monthly 

 tables of diurnal inequality, not from combinations of diurnal inequalities 

 for longer periods. 



Table XII. contains for Greenwich the results of a determination of 

 the difference between the absolute values of declination and horizontal 

 and vertical force as found from all days and from quiet days in the years 

 1889 to 1896, combining the same month in different years for monthly 

 differences, and giving also differences for separate years. The quiet day 

 mean was found by selecting, for the several quiet days in each month, 

 the daily means in declination and horizontal and vertical force from 

 Tables I., III., and VII. of the Greenwich annual volumes, and comparing 

 the respective monthly means of the same with the monthly means for all 

 days taken from the Greenwich Table XI. Our table thus gives differences 

 only — the excess of the quiet day value over the all day value. In addition 

 to the mean excess of absolute value, there is given for each element also 

 the greatest and least monthly values of excess, both for the monthly 

 grouping and the yearly grouping. Tlie mean excess in declination shows 

 that the quiet day value was slightly greater than the all day value in 

 nine months of the year, although the numbers in the column ' Least 

 monthly excess ' show that in each month the quiet day value was in some 

 years in defect. In horizontal force the mean excess shows that the quiet 

 day value was greater than the all day value in all months — in some con- 

 siderably so — although in eleven months of the year the quiet day value 

 was in some years in defect. In vertical force the mean excess shows that 

 the quiet day value was less than the all day value in nine months of the 

 year, although in each month the quiet day value was in some years in 

 excess. The mean of the monthly numbers in the column ' Difference ' 

 is less than that of the yearly numbers, in all elements, as was found under 

 similar conditions in Tables VIII. and XI. 



In regard to the circumstance that the mean day of the five selected 

 quiet days does not always fall exactly at the middle of the month, and 

 the influence that this may have on the quiet day value, and consequently 

 on the numbers of Table XII., it is to be remarked that in the monthly 

 grouping eight months of different years are combined, and in the yearly 

 grouping of course the twelve months of each year. And in neither of 

 these groupings does the mean day resulting from the combination of 

 months in any case differ by more than one day and a quarter from the 

 middle of the month. Taking the annual decrease of declination as 6'0, 



