ye 
has since discussed the two-hourly observations made with the 
bifilar magnetometer in the years 1841, 1842, 1843, the whole 
series being re-arranged according to the moon’s hour-angle, 
in the manner already described in the corresponding inves- 
tigation relating to the declination. No correction has been 
applied for temperature, the effect of the diurnal variation of 
temperature being assumed to be eliminated in this mode of 
grouping the results. 
The following Table contains the yearly mean results for 
the several lunar hours, reckoned from the time of the moon’s 
upper meridian passage. The numbers are the differences 
between the horizontal intensity at each hour, and that of the 
entire day, expressed in millionths of the intensity. The re- 
sults are given for each year separately, and for the mean of 
the three :— 
TaBiE I.—Diurnal Variation of the Horizontal Intensity related to the 
Moon’s Hour-Angle. Yearly Means. 
Hours. 1841. 1842. 1843. Mean. 
0 = 45 ee” + 34 - 16 
2 — 130 — 94 - 9 - 78 
4 - 57 0 -— 18 ie 
6 + 6 + 64 -1l + 19 
8 +173 +94 + 55 + 107 
10 +116 + 80 + 30 a iG 
12 + 80 - 4 - 2 + 25 
14 + 62 — 50 — 32 sy, 
16 — 100 — 4] — 82 =a 
18 = N8D — 32 —14 — 42 
20 + 47 — 25 + 55 + 12 
22 — 27 + 9 -— 4 ai G, 
It appears from the foregoing Table that the horizontal 
component of the magnetic intensity is subject to a periodical 
variation in the course of the lunar day, analogous to that 
already established in the case of the declination. The hori- 
