554 
MR. W. ELLIS ON THE RELATION BETWEEN MAGNETIC DIURNAL 
In laying down on Plate 23 the numbers of the foregoing table, the same relative 
scale has been employed as on Plate 22, that is to say, one minute of arc of declination 
is taken to correspond to 20 - 0 in sun-spot number. 
The appearance of Plate 23 is very different from that of Plate 22, the special pecu¬ 
liarity of each month being now fully displayed; at the same time the general eleven 
year relation is also distinctly apparent. In regard to -what may be called minor 
variations, the correspondence between the curves is not always of a marked character ; 
but in some of the greater and more sudden manifestations of energy, the agreement 
is very striking. Thus the sudden increase of sun-spot activity in the middle of the 
year 1847 is accompanied by a no less sudden rise in the declination curve, and, in 
both cases also, the increased activity is for some time maintained. But no corres¬ 
ponding motion of similar extent is to be seen in the horizontal force trace, although 
there is a sudden increase, previously, in 1846, and another, afterwards, in 1848, the 
former of which nearly agrees with a lesser upward movement in the sun-spot curve, 
and the latter with renewal of activity in the same curve. Again, various correspon¬ 
dences near the epoch of the 1870 maximum are very remarkable; in each of the 
years 1869, 1870, 1871, and 1872 there are upward motions, the counterparts of which 
even to some of the smaller bends are to be seen in all three curves. A sudden fall 
in the year 1873, without upward return, is also shown both in the sun-spot and 
declination curves, although hardly with equal distinctness in the horizontal force curve. 
In 1869 the highest point in the sun-spot curve is reached in June, the highest points 
in the two magnetic curves being reached in July and June respectively; in 1870 the 
highest point of the sun-spot curve is reached in May, and the corresponding points in 
the two magnetic curves in July and June respectively; in 1871 the highest point in 
the sun-spot curve is reached in April, the highest points in the two magnetic curves 
both occurring also in April. Generally, the variations about the period of the 1870 
maximum occur so nearly simultaneously in the three curves that it does not definitely 
appear that there is any real difference of epoch. The results deduced in Tables III. 
and IV., from consideration of the epochs of minimum and maximum only, have pre¬ 
viously shown the difference to be small. The presumption, in regard to epoch, is that 
if the various phases of sun-spot and magnetic effect are not entirely coincident, the 
latter follow the former by comparatively short intervals of time. 
It seems worth pointing out that at each of the three epochs of maximum the sun¬ 
spot curve exhibits a double maximum ; the similarity of the manifestation in the 
first and third cases, 1848 and 1870, being remarkable. The three curves show in 
general a much closer agreement during the later years, which seems to confirm the 
impression produced by the consideration of Plate 22, that a greater accuracy in the 
magnetic indications has been attained since the instruments have been located in the 
magnetic basement, that is since the beginning of the year 1865. 
A further examination of Plate 23 shows that although the average annual inequality 
in the magnetic curves has been removed, there yet remains, in some years, a very 
