ATMOSPHERIC MAGNETISM — DIURNAL VARIATION. 
69 
atmosphere, spoken of the resultant in the atmosphere dependent on the sun’s pre- 
sence ; and will do so a little while longer without implying any direct action of this 
resultant, or that portion of air which yields it, upon the needle (2933.), for the sake of 
considering at what probable height it is situated in the air. That it cannot be on the 
surface of the earth, is shown by the depression of the lines and diminution of the dip 
at St. Helena and Singapore during the middle of the day ; and that it is not even 
under the sun, is shown by the manner in which the greatest action precedes, in some 
degree, the sun, as at Hobarton and Toronto, and other places by different amounts 
of time ; neither the time when the sun is on the meridian, nor the time when the 
observed temperature is highest (for that is after the sun), is the time of greatest 
action, but one before either of these periods. The changes in the temperature of the 
air produced by the sun, will not take place below and above at the same time. 
The upper regions of the atmosphere over a given spot are affected by the sun at his 
rising and afterwards, before the air below is heated; and therefore the effect from 
above would be expected to precede that below. The temperature observed on the 
earth does not show us, for the same time, the course of the changes above, and may 
be a very imperfect indication of them. The maximum temperature below is often 
two, three, or four hours after the sun, whereas, whatever heat the sun gives by his 
rays directly to the atmosphere, must be acquired far more rapidly than that. It is 
very probable, and almost certain, that at 4 or 5 o’clock a.m. in the summer months, 
the upper regions may be rising in temperature, whilst on the surface of the earth, 
through radiation and other causes, it is falling. The well-known effect of cold 
just before sunrise in some parts of India, and even in our country, is in favour of 
such a supposition. We must remember that it is not the absolute temperature of 
the air at any spot that renders it influential in producing magnetic variations, but 
the differences of temperature between it and surrounding regions. Though the 
upper regions be colder than the lower, their changes may be as great or greater ; 
they happen at a range of temperature which is probably more influential than a 
higher range (2967.) ; and, what is of importance, they occur more quickly and 
directly upon the presence of the sun. The quantity of heat which the atmosphere 
can take directly from the sun’s rays, is indicated by the different proportions we 
receive from him when he is either vertical or oblique to us, and so sending his 
beajus through less or more air ; and when he has departed, the upper parts of the air 
are far more favourably circumstanced for rapid cooling by radiation than the por- 
tions below. So that the final changes may be as great or greater than below, and 
we may learn little of them, or their order, or time, by observations of temperature 
at the earth’s surface. In addition therefore to observations of magnetic effect, as 
depression of the lines of force at St. Helena, &c., there are apparently reasons de- 
ducible from physical causes, why the chief seat of action should be above in the 
atmosphere. 
2938. In the midday effect the upper end of the needle passes the mean position 
