ATMOSPHERIC MAGNETISM — ST. PETERSBURGH VARIATIONS. 
99 
come on the eastening from the cold ceases sooner and sooner, being in January and 
April 17 ^ and IS** respectively. The minimum of temperature also retreats, being for 
the same month 20^^ and 16^*. On the contrary, the maximum of heat advances from 
the winter to the summer months, being also greatly increased ; and the effect on the 
sun-swing is seen both in the advanced time of change and the increased amount of 
variation. 
3015. In May and June the night or cold eastening has disappeared, or is shown 
only by a little hesitation ; and from midnight the coming on of the sun region sets 
the needle end west. If we look at the globe (2996.), we should be led to expect that 
it would do this. The sun is then in the northern tropic nearly, wheeling round 
St. Petersburgh and comparatively near to it; and a free dipping-needle would in 
twenty-four hours make one revolution in the same direction as the sun region, but 
at the opposite end of the line, joining the two together. If the needle were at the 
astronomical pole of the earth, having great dip, it would describe almost a circle 
with nearly uniform motion ; but being really much nearer to the warm region in one 
part of the uniform daily course of the latter than another, the radius vector joining it 
with the region then makes a much greater angle in a given time than when it is fur- 
ther off, and hence the greater rapidity of the motion between 20'‘ and 1^, and the 
production of what I have familiarly called the sun-swing from west to east. 
3016. It will be seen from the Table of Curves (Plate II.), that we have at St. Peters- 
burgh a fine example of that kind of result which Colonel Sabine called attention to 
so strongly in his paper upon the St. Helena phenomena* ; and those occurring at 
Hobarton, Toronto and elsewhere ; namely, a declination variation in different direc- 
tions for the same hours in different months. Thus, in the present case, the needle 
end goes eastward for the hours 13^ to 20^ in October, November, December, January 
and February, whilst it goes west for the same hours in April, May, June, July and 
August : March and September curves fall midway. But this difference is now I hope 
by the hypothesis accounted for (3010. 3015.), and I trust that equally satisfactory 
reasons will appear for St. Helena (3045.) and other places (3022. 3039. 3065.). 
3017 . The paramagnetic character of the eastening effect by cold in the winter 
months after 10 o’clock, would probably be illustrated by inclination observations for 
the same time ; for if the cold region passes to the south of St. Petersburgh the inclina- 
tion will be decreased by the paramagnetic action, but increased by the diamagnetic 
resultant ; and the manner in which these two elements of direction, i. e. inclination 
and declination, are combined at any given moment, is very important to the full 
elucidation of the magnetic effect of the atmosphere. I have not been able to give 
these data for St. Petersburgh. The total force variations would also help greatly to 
clear up the subject. Indeed it is not fair to endeavour to explain the results of the 
assigned cause by taking only one element of three into consideration. What we re- 
quire ultimately to know, is all the changes of a free needle in position and in respect 
* Philosophical Transactions, 1847, p. 51. 
o 2 
