ATMOSPHERIC MAGNETISM— SINGAPORE. 
113 
therefore deflects the lines of force upwards and increases the dip, and so the portions 
of ellipse at the two places, which correspond in time and direction as to declination, 
have contrary variations of inclination. 
3058. Singapore^. — This is a very interesting station ; being in longitude 103° 53' E., 
it has only 1° 16' N. latitude, and so is close to the equator. Its declination too is 
only 1°40'E., and its inclination 12° S. It is also near the line of weakest force round 
the earth. The magnetic equator of the needle is almost parallel to the earth’s equator, 
and the quadrants (2929.) are distributed with great simplicity, the magnetic and 
astronomical meridian nearly coinciding. In our summer the sun passes through 
the east and west northern quadrants during the daytime ; in our winter through the 
east and west southern quadrants ; and in certain months through all four quadrants, 
following nearly the neutral line of the magnetic equator. 
3059. Hence if the line of force were free, i. e. if it had no hold in the earth (2919.'), 
we should expect from the hypothesis little or no change in the needle, especially in 
the months when the sun was over the magnetic equator ; but because there is dip, 
and the lines of force which govern the needle are to the south tied up in the 
earth (2929.), whilst they are free to move in the air and space toward the north, so 
there is variation both of the declination and inclination in a perfectly consistent 
manner ; and keeping this in mind, I think we shall have no difficulty in tracing the 
monthly results according to the hypothesis. 
3060. In the first place, the curves of day variation are so like those of St. Helena, 
month for month, that the account given of them there will suffice for the present 
occasion (3048.). The sun-swing occurs at the same period, and the effect, dependent, 
as I suppose, on the character of the two hemispheres, is produced (3054. 2949.). There 
are however striking differences in the latter part of the sun turn, and also in the 
night hours, from 5^ to 14’\ The amount of variation appears small ; but this is chiefly 
due to the circumstance that the horizontal plane on which we read it, almost coin- 
cides with the free needle, and so the correction before referred to (3009. note) neces- 
sary to give the true value- of the variation is here very small. 
3061. Considering June first, as at St. Helena, the upper needle end moves east as 
before until 19**, under the influence of the morning cold, after which it stops and is 
sun-driven west until 22**, when it swings downwards and beneath to east by 3** ; then 
follows the sun west until 7^ ; it then stops and returns, creeping more and more east 
because of the coming cold (3065.). In July the needle eastens a little more before 
19**; westens until 23**, and then eastens until 4*’. The sun-swing is thus thrown an 
hour later than in June, which I believe to be connected with the accumulation of 
heat over the land (3054.), combined with the lagging effect of the sun (3053.). In 
August the needle end eastens until 19**; more than in July, and most of all the 
months : it then westens strongly before the sun until 23**, after which the sun-swing 
* See Tables, pp. 121, 122, and the curves of variation, Plate II. The data for Singapore are deduced from 
the recent very valuable labours of Captain Elliot. 
MDCCCLI. Q 
