Annual Variations for the Vertical Component of Magnetic Force, xlvii 
Table 41. — Mean Change of the Value of the Vertical Component from Month to Month for 
diffei'ent Groups of Years, as deduced from Table 38. 
Dec. 
Jan. 
Feb. 
March 
April 
May 
.June 
July 
Aug. 
Sept. 
Oct. 
Nov. 
Yeai's. 
to 
to 
to 
to 
to 
to 
to 
to 
to 
to 
to 
to 
Mean. 
Jan. 
Feb. 
March. 
ApriL 
May. 
June. 
July. 
Aug. 
Sept. 
Oct. 
Nov. 
Dec. 
Prefix. 
0-aiio 
0-000 
0-000 
0-000 
0-000 
0-000 
0-000 
0-000 
0-000 
0-OUO 
0-000 
0-000 
0-000 
1843-6 
-219 
-261 
- 178 
- 135 
-150 
-f 008 
-208 
- 160 
- 170 
-190 
-078 
- 129 
- 156 
1842-9 
-205 
-166 
-194 
-137 
- 144 
-048 
-090 
- 194 
-161 
-019 
-049 
- 128 
- 128 
1842-7 
- 194 
-183 
-197 
-097 
- 132 
-039 
- 196 
- 159 
-141 
-134 
-064 
- 153 
- 141 
88. Considering the numbers for the years 1843-6, we find that the mean change of the vertical component 
from one month to the next = 0-0001 5 6, that the diminution in the months from December to Mai'ch and from 
June to October was greater than the mean, while those from Marcli to June and from October to December 
were less ; the other groups give nearly the same result, which is quite in accordance with that from Table 40. 
From both Tables we feel entitled to state the following as the annual law, — That the vertical component of 
magnetic force is a maximum near the solstices and a minimum near the equinoxes. It will be observed that 
this is precisely the law already deduced for the horizontal component No. 56 ; had it not been for this remark- 
able coincidence in a law with two maxima and two minima, obtained from two instruments of the most different 
principles, the conclusions deduced from the observations of the balance magnet would have been left with 
whatever weight they might appear to physicists to deserve ; but it is conceived that the agreement is too con- 
siderable and too remarkable to omit adducing it as evidence in estimating the accuracy of this result for 
the vertical component. 
89. If we compare the monthly means deduced from the days selected in each month as nearly free from in- 
termittent disturbance, with those deduced from all the hovirly observations in the corresponding months, we 
find the latter less ( — ) or greater ( + ) by the following quantities. 
Year. Prefix. Jan. Feb. March. April. May. June. July. Aug. Sept. Oct. Nov. Dec. 
1844, 0-000 -039 +096 -124 +003 -071 -031 -031 -020 -048 -043 +006 +051 
1845, 0-000 -068 -036 -039 -086 -061 +018 -049 -091 -049 -046 +008 +024 
Mean, 0-000 -053 +030 -081 -041 -066 -006 -040 -055 -048 -044 +00? +037 
The numbers differ considerably in some cases for the same month in the two years ; a greater number 
of partial results are therefore evidently required for a good mean. The mean of both years shews, that the dis- 
turbed means were greater than the undisturbed in November, December, and February (or, about the winter 
solstice), and less in all the other months, the diminution being greatest in March, May, and August. 
Differences of the Daily Means of the Vertical Component from the Means for the corresponding Months. — 
The discussion for 1844 will be found in the volume for that year, p. 374, the results for 1845 and 1846 
are obtained from Tables XXXI. and LXIIL, pages 18 and 35 of this volume. 
90. The conclusions fi-om Table 42 are : — 
\st, That the positive departures of the daily mean vertical component from the monthly mean value are 
greatest in September and in February, and that they are least in January and June. 
2d, That the negative departures of the daily mean fi'om the monthly mean are greatest in September, 
January, and May, and least in March and July. 
3(i, That the mean positive departure is most in excess of the mean negative departure in February, while 
the latter is most in excess of the former in January, and in the montlis from March to June ; with the excep- 
tion of February, the mean negative departure is greater than the mean positive departure in the first six months 
of the year ; and, with the exception of August, the reverse is the case for the last six months. 
4:th, That the mean departure of the daily mean from the monthly mean (without reference to sign) is 
greatest in September, and least in March. 
5*A, The mean departure of the daily mean vertical component from the monthly mean for 1844 = 0-000105 
, 1845 = 0-000117 
1846 = 0-000131 
3 years = 0-000118 
MAG. AND MET. OBS. 1845 AND 1846. 
m 
