RANGE AND THE PERIOD OF SOLAR SPOT FREQUENCY. 
545 
The observations were partly interrupted during the year 1864 in consequence of 
the construction of the “magnetic basement,” to be alluded to hereafter. The 
values for 1864, inserted in the preceding table, both for declination and horizontal 
force, are inferred values, as also are those for horizontal force for January 1847, July 
1861, and January 1865. All other numbers (excepting those for the years 1865, 
1866, and 1867, which have only recently been deduced from the photographs) may 
be verified by reference to the several annual volumes of ‘ Greenwich Magnetical and 
Meteorological Observations/ 1841 to 1847, to the ‘ Results of Magnetical and Meteoro¬ 
logical Observations ’for 1859 and 1867, and to those for the several years commencing 
with 1868. The numbers for the years 1841 to 1847 will, in some cases, slightly differ 
from those to be obtained from the several printed volumes, because, in the formation 
of the magnetic abstracts, until the year 1847, no separation of days of unusual 
magnetic disturbance was made. And commencing with the year 1868 the numbers 
for horizontal force, as given by the yearly volumes, require correction for temperature: 
the correction is, however, very small, and has been here duly applied. 
The increase of the numbers in the summer months in both elements is, in the 
preceding table, plainly apparent. But in order to estimate progressive change, a 
number for each month is required which shall be free of annual inequality, and such 
number has been formed as follows. Taking, for example, the month of July, the 
new number—suppose for declination—is equal to 
-^[Number for January (preceding) + Number for January (following) ] 
-f ^(Number for February+Number for March.-j-Number for December) 
and it represents an annual mean applying to the year whose centre is the middle of 
July. And similarly for each individual month. The process, which assumes the 
months to be equal in length, is equivalent to taking the means of each twelve 
consecutive monthly numbers, and again taking the means of each two consecutive 
numbers. Thus is obtained, both for declination and horizontal force, a set of 
numbers practically free of annual inequality. Throughout this discussion the effect 
of lunar inequalities, as being presumably small, is neglected. The new numbers are 
contained in the following table. 
