ANNUAL VARIATIONS FOR THE MAGNETIC DECLINATION. Xi 
TABLE 3.—Monthly Variations of Magnetic Declination free from Regular Secular Change. 
1847, | 
eae ce 8 Corrected by | ae 
Month. and M 1848. 1849. 3 Years, 
Maker- Green- | | 1847-9. 
stoun. | wich. | 
| = 
, , | ’ , ‘ 
January . . . -38 |—0-38 |—0-44|40-51| —0-34 | —0-08 
February . . . 07 | —0- . -02 |+0-05 |+0-33 |—0-13| +0-06 ||+0-04 
March 5 : : p | . . : + 0-74/+0-75|+0-68) +0-24 || +0-55 
April . . . : . -44 [4+1-24/+1-17|+0-31| —0.25 | +0-42 
May . . . : 6 E 0-00 }+0-18 | +0-42) +0-21 | +0-24 
June . . . . . . -14 |—0-28 |—0-20|—0-69| +0-41 |—0-17 
July : : . + 0- -30 [40-18 |40-10|+0-09| —0-74 | —0-17 
August 05 | +0-65 : -45 |+0-40 |4 0-27 |—0-32| —0-60 || —0-20 
September|| — 0- . ce : -38 |—0-88 |—1-03 |—1-15 | —1-01 | —1-04 
October “Ti : +25 : . 55 |—0-83 |—0-25 |—0-38| +0-10 | —0-27 
November . . . * : : : — 0-01 |—0-48 —0-19)| +0-92 ||+0-16 
December : /—0- c 41 |—0.28 |-0-36 +0-84 [41.06] +0-53 
8. Table 3 has been formed from Table 1 in the following manner ; the monthly means for 18438, 4, 5, and 6 
were reduced for mean secular change to January of their respective years by the correction 
M, + 0°48 x n 
where M,, is the mean for the xt month after January, and 0/48 is the approximate mean value of secular 
change for one month. If m be the mean of the twelve resulting quantities for any year, the numbers w in 
Table 3 are obtained by the formula 
(én = M, + 0°48 x n—m 
The numbers for 1847, 8, and 9, were obtained in a similar manner; 0'-63 being used instead of 0'-48 
for 1847, 0'-62 for 1848, and 0'-56 for 1849 ; the value of the secular change for 1847 has been obtained by 
comparing the last six months of 1846 with the corresponding months of 1847, and the first six of 1847 with 
the corresponding months of 1848 ; that for 1848 was obtained similarly. The means for 1841-2 are not in- 
serted, as they were too much affected by torsion and broken suspension-threads to be of use in this investigation. 
9. The interpolated epochs of maximum and minimum, from the mean of 4 years in column 8 of Table 3, 
are,— 
A minimum of westerly declination in the end of April ; 
AsMaximuMiipgereescnescsc serves in September ; 
Al nin imu peeseeee een «en eecise0: in the beginning of December ; 
Al maximum reese cence cscs in the end of January. 
This result is shewn with considerable fidelity in each of the four years; the greatest variations from it 
can be traced to torsion of the suspension thread removed at the particular epochs: the means for 1846 give 
accurately the result of the means for the other three years. 
10. The year 1847 was one of great magnetic disturbance, and as only 5 observations were made daily, the 
effect of the disturbed observations on the monthly means is the more considerable. Thinking it possible that 
corrections for the 5 observations might be obtained with greater accuracy from complete series made elsewhere 
during the same year, I applied to Mr Arry, the Astronomer-Royal, for this end. I have to thank him for fur- 
nishing me with corrections obtained from the Greenwich Observations for that year. These corrections having 
been applied, the resulting variations, obtained as previously indicated, will be found column 10 of Table 3; 
they give almost exactly the same result as the quantities corrected by the Makerstoun Observations for 1844 
and 1845: according to both, there is a slight maximum exhibited in August, but otherwise the result 
differs considerably from that given by the preceding years. The observations for 1848 and 1849 on the 
whole indicate a result not differing greatly from that for 1847; and when we examine the mean for the 
3 years 1847-9, as in the last column of Table 3 (where the mean of columns 9 and 10 has been taken for 1847 ): 
we find a similar but more regular result. The value of the conclusions from the observations for 1843-6 
depends upon the consistency of the partial results and the regularity of the secular motion from year to year : 
when it is pointed out that the means for 1847, 8, and 9, are deduced from but few daily observations, it should 
