ATMOSPHERIC MAGNETISM — DIURNAL VARIATION — DIRECTION. 
63 
mean declination 38° E. These observations are only for two months,?, e. April and 
May 1844. The extreme western place of the upper or south end of the needle was 
at 19 o’clock, and its extreme eastern position at 2 o’clock. The result therefore is 
in perfect accordance with the preceding observations and conclusions. The amount 
of variation, as given in the horizontal plane, is very large, being 36''26 in April and 
32' in May. 
2915. St. Petershurgh . — Latitude 59° 57' N. ; longitude 30° 15' E. of Greenwich ; 
mean declination 6° 10' W. ; the dip 70° 30' N. The observations are the mean of six 
years, and show that the upper end of the needle is extreme west in regard of noon, 
about 19*^ and 20*' for the months of March to August, and that for the other 
months there is a western position about the same hours. The extreme east position 
is, for ail the months, about half-past 1 o’clock, so that the sun’s effect in passing- 
over at the noon period is as in former cases. The greatest amount of variation is 
ll'*52 in June; in- winter it dwindles away to as little as 1'‘77- From theory, the dip 
may be expected to increase during the day hours and diminish at night. 
2916. Thus these cases, which, including the chief feature of diurnal variation and 
sun action, were selected as a first and trial-test of the hypothesis, join their evidence 
together, as far as they go, in favour of that view which I am offering for their cause ; 
nor have I yet found any instance of even an apparent contradiction in regard to the 
sun action. They assist the mind gieatly in forming a precise notion of the manner 
in which the influence of the sun and air is supposed to act, not only in similar cases, 
but in respect of other consequences, i. e. in all that properly comes under the term of 
atmospheric magnetism ; I will therefore now restate more particularly the principles 
which, according to the hypothesis, govern them, in hopes that 1 may be fortunate 
enough to assist in developing by degrees the true 'physical cause of the magnetic 
variations in question, 
2917- Space, void of matter, admits of the transmission of the magnetic force 
through it (2787. 2851.). Paramagnetic and diamagnetic bodies either increase or 
diminish the degree in which the transmission takes place (2789.). This, their in- 
fluence, I have expressed, for the time, by the phrase of magnetic conducting power, 
and I think have given suflUcient first experimental evidence of the existence of the 
power and its effects in disturbing the lines of magnetic force (2843.). The atmo- 
sphere is, by the oxygen it contains (2861. 2863.), a paramagnetic medium, and has 
its conducting force greatly diminished by elevation of temperature (2856.) and by 
rarefaction (2782. 2783.), as has also been fully proved by experiment. The sun is 
an agent which both heats and rarefies the atmosphere, and in its diurnal course, the 
place of greatest heat and rarefaction must, speaking generally, be beneath it. The 
irregularities in the condition of the earth’s surface and other causes do produce local 
departures from an exact relation of place, but they probably disappear partly, if not 
altogether, in the upper regions of the air. 
2918. Assuming that the air under the sun is most changed magnetically, and con- 
