114 DR. FARADAY’S EXPERIMENTAL RESEARCHES IN ELECTRICITY. (SERIES XXVII.) 
comes on and continues until 5**, as if the warm region were behind the sun, perhaps 
even 2'". The time of the swing is much prolonged, and not unnaturally, as the 
place is at the equator and therefore under the sun. In September the eastening is 
less, the westening is less, and the sun-swing is less. April is like September, except 
that the latter shows the effect of the previously warmed hemisphere (3053.). 
3062. Then there are four months in the year, November, December, January and 
February, when the sun is south of Singapore, and altogether during the day in the 
southern quadrants (3058.). As the sun comes on from IG'' or 17^ the upper part of the 
line of force moves westward (the lower being fixed in the earth) until 19 or 20 o’clock. 
The sun is at this time in the south-east quadrant, and it might be expected perhaps 
that the motion of the north or upper end of the needle should be to the east if there 
were any change at all. But there are two or three reasons, from the hypothesis, why 
this should not be. For that effect there should first of all be no dip ; and in the next 
place, if there were no dip, the sun is so nearly in the neutral line of the magnetic 
equator, that the deflection, if any, would have been very small. On the other hand, 
the lines of force have dip to the south, and being therefore held in the earth ; that 
travelling of the sun along the neutral line, which in its coming up would have sent 
the whole line of force west, and so caused no variation of declination, can now only 
send the northern parts, as they rise out of the earth and are carried on with the 
general system of lines, west, and so cause that western travelling of the needle which 
does occur. Besides this, though the sun be south of that neutral line and also of 
Singapore, there is reason to suppose that the middle or resultant of the warm region 
is north even of both (3063.), which would aid the westening of the needle just de- 
scribed. 
3063. For if we recall to mind Dove’s results, they show that the northern hemi- 
sphere, as a whole, is warmer than the southern (2949.). Again, if we look at the 
meridian of Singapore, we shall find that there is far more continent on the north of 
it, to produce a higher temperature, than to the south ; and even by the local tables 
of temperature below, we shall find that May, June, July and August are the hottest 
months for Singapore, and November, December, January and February the coldest ; 
all tending to make us suppose that the warm region of the atmosphere is relatively 
north of the sun’s place, and perhaps even of Singapore (3067.)* 
3064. At 20^ the sun-swing from west to east comes on, and continues until 2 ^, 
after which the needle moves west, following the sun, until lO** or iD when it is near 
the mean ; it still goes on westening very slowly until 17 ^ when the morning sun action 
takes it up and drives it more quickly west, until about 20’", when the sun-swing east 
occurs. The curve in these months is very simple in its character ; the night or cold 
effect appears to be but small, being indicated rather by a hesitation than by a 
distinct movement east. 
3065. The easterly movement of the needle end in May, June, July and August, 
and the westerly movement in November, December, January and February, for the 
