70 DR. FARADAY’S EXPERIMENTAL RESEARCHES IN ELECTRICITY. (SERIES XXVI.) 
(2935.) on its return to the east generally before the sun passes the meridian going 
westward. At Toronto it is about an hour in advance ; at St. Helena and Wash- 
ington an hour and a half ; at Greenwich and Petersburgh two hours ; at Hobarton 
and the Cape of Good Hope the passage is about noon. Such results appear to in- 
dicate that the place of maximum action is in advance of the sun ; and it probably is 
so in some degree, but not so much as at first may be supposed, as will appear 1 
think from the following considerations. 
2939. The precession of the time of maximum action may depend in part upon some 
such condition as the following. As the sun advances towards and passes over a meri- 
dian, the air is first raised in temperature and then allowed to fall, and these actions 
produce the differences in different places on which the magnetic variations depend. 
But they depend also upon the suddenness with which or the vicinity at which these 
differences occur. Thus two masses of air, having equal differences of temperature, 
will affect the lines of force more if they be near together, and to the needle, than if 
they be far apart. And again, if a body of air were of a certain low temperature at 
one part, and, proceeding horizontally, were to increase rapidly to a certain high tem- 
perature and then diminish slowly to the first low temperature, such a body passing 
across a set of lines of magnetic force would affect them in opposite directions at the 
fore and after part ; but it would affect them most on the rapidly altering side. 
2940. Now the air as heated by the sun must be in this condition. According to 
analogy with solid and liquid bodies, being exposed to heat and then withdrawn, 
the changes of temperature that it would undergo would be more rapid in the eleva- 
tion than in the falling, and so the changes in the preceding would be more rapid than 
in the following parts. To this would be added the effect of the atmosphere warmed 
by the earth ; for as that is slower in attaining heat, as is shown by the time of maxi- 
mum temperature, so its effects being gradually communicated to the air above, as 
the sun passed away, would tend to retard its fall and enlarge the difference already 
spoken of. Applying these considerations to the natural case, the strongest effect 
and the greatest variation should be towards the west, and the following or lesser 
action towards the east of the sun ; and the mean condition of the needle for the 
whole change would be in advance of that body. 
2941. Mr. Broun has made observations of the daily variation at different heights, 
namely, at Makerstoun and the top of the Cheviot Hills, where the height differs by 
nearly half a mile, and finds, I believe, no difference in the intensity, but that the 
progress is Jirst at the higher station. It would be very interesting to have an ob- 
servatory up above, but to give the results required it should have air and not solid 
matter beneath it. 
2942. There is another circumstance which importantly influences the times of the 
passages of the declination variation. If two places north and south of the equator 
have equal dip and contrary declinations, i. e. if both their upper ends point east or west, 
then the effects ought to correspond and form a pair. But if both have east or west 
