XX GENERAL RESULTS OF THE MAKERSTOUN OBSERVATIONS. 
lst, The departure of the declination magnet, at any hour, from the monthly mean position, for the same 
hour, is greatest two or three days after opposition, and it is least about the time of conjunction. 
2d, The departure of the declination magnet, at any hour, from its monthly mean position, for the same 
hour, is least when the moon is farthest north, a minimum also occurring after the greatest southerly declina- 
tion ; it is greatest about 4 days after the moon is farthest north, a maximum also occurring about 4 days before 
that ‘epoch. 
25. There are slight irregularities in the resulting means of Table 11, which are to be expected where only two 
years’ observations are considered. The variations of these differences are rather less for the period compre- 
hending the moon’s changes of declination, than for that comprehending the sun’s ; the means for 2 or 3 days 
of the former being compared with those for the 12 months in the latter; if the mean difference for June, 
however, be neglected, the variations for the lunar period are quite as great as those for the solar period. 
TABLE 11.—Mean Difference of a Single Observation of Magnetic Declination from the Monthly Mean, 
for the corresponding hour, with reference to the Moon’s Age and Declination. 
i | After 
Moon’s | Varia- Moon 
tions. farthest 
North. 
d. da. | 
o7=—= 11 
2— 5 
6— 8 
9—12 
V3—15 
16—19 || 
20—22 | 
23—26 | 
tH} 
DIURNAL VARIATIONS FOR THE MAGNETIC DECLINATION. 
26. Diurnal Variation of Westerly Declination.—The discussions for 1843 and 1844 will be found in the 
volumes for these years; the tables for 1845 and 1846 are contained in this volume, pages 4 and 31. Table 
12 is formed from a combination of all the ordinary daily observations made in these years in the following 
manner. Let A and C be the means from the 4 years’ observations for two hours at 7 hours interval, the in- 
termediate hours having less than 4 years’ observations, « and ¢ the means from the observations for 1844 and 
1845 at the corresponding hours ; b, the mean for the nt* hour after a, for which there are less than 4 years’ 
observations ; then B,, the mean for the same hour referred to the means A and C has been obtained by the 
formula 
i 
————s ——— n 
B,— bn = (c—<_a=0) a 
where b’, = b, — a — A. 
. Between 17" 10™ and 92 10™, Makerstoun mean time, only the even hours have not had observations 
for ie ee ; for these even hours, therefore, i i = 2, n = 1, and the formula is reduced to 
A—a+C—ec 
a 
2 
For 225 10™, 0" 10™, and 6h 10™, b depends on 3 years’ observations, 1844, 1845, and 1846, and so there- 
fore do @ and ¢ in the formula for these hours ; and in November and December the mean for 22" 10™ jg de- 
duced from 4 years’ observations (see foot-note, p. xi.) This reduction is evidently the simplest and least exact- 
ing that could be made. 
