ATMOSPHERIC MAGNETISM — DIURNAL VARIATION — DIRECTION. 
59 
2901. All this may be roughly represented by fig. 1 1 (2909 ), in which E.W. represents 
the path of the sun between the tropics as he comes up with the hours 21'*, 22'*, &c. 
in his daily journey, and e the path described by the north or upper end of the needle, 
freely suspended at Hobarton, and therefore showing both declination and inclina- 
tion, i. e. the whole direction. Looking down upon such a needle, its upper end will 
take the course indicated by the arrow, and its position at any given hour is shown 
sufficiently by the leading lines. 
2902. This relation of the motion of the needle to that of the sun has long been 
known; it has great significance in relation to my hypothesis of the physical cause 
of these variations. As regards the part of the action which I am considering, it is 
as if the pole of a magnet came on with the sun, of like nature to the upper end of 
the Hobarton needle, and at first drives that end west. Towards 19 o’clock the 
tendency westward diminishes, but the tendency south increases. At 21 o’clock, the 
increase in the sun’s power, acting not directly from the sun but from a region in the 
atmosphere beneath it, is not sufficient to compensate for his more unfavourable 
position ; the earth’s force brings the needle back as regards declination, and then 
it passes eastwards, but the southerly motion or inclination still increases ; about 
24 o’clock, or noon, the sun is as to east or west declination indifferent, but powerful 
in southern action, making the inclination then, or soon after, a maximum. Then as 
the sun goes west of the needle, its power in driving the pole behind it eastward, will 
increase for a time, whilst the power producing inclination will diminish, until at 2 or 
3 o’clock the earth’s force will regain preponderance as the sun’s power diminishes by 
distance, and the needle will return towards its least dip and mean inclination. 
2903. All this may be represented experimentally by carrying a magnetic pole 
north of the dipping-needle, so as to represent the place of the sun-heated air to 
Hobarton, provided that pole be of the same kind as the north or upper pole of the 
needle. I have already stated (2877- 2863.), that when a portion of air is heated in a 
field of magnetic power, it loses in magnetic conduction power, and if in associa- 
tion with air less heated deflects the lines, assuming the state which I have distin- 
guished as that of diamagnetic conduction polarity ; then presenting the very polarity, 
or rather the very inflection of the lines of force, which would affect the needle, as it 
is affected. As the sun rises and passes north of such a place as Hobarton, the atmo- 
sphere under his coming influence becomes more and more heated and expanded ; 
and referring to the model globes of air (2864. 2877-)? if 'S as if such a warm mass 
passed with the sun through all the regions of the equator, extending also far north 
and south of it ; and having Hobarton within its influence, produced the effects there 
observed. 
2904. In such a view one sees a reason for the short time occupied in the return 
of the needle from west to east as the sun passes immediately over its meridian, and 
for the long time during which it is passing from east to west as the influence of the 
I 2 
