ATMOSPHERIC MAGNETISM — ST. HELENA. 
109 
of November, December, January and February, or for nearly that time; it is north 
of the island for the rest of the year. At one time the sun passes the astronomical 
meridian before it arrives at the magnetic meridian, and at another time the con- 
trary is the case. In addition to these peculiar circumstances, St. Helena is a place 
of great local differences, and also its dip is so low that the sun’s day effect is almost 
constantly to depress and lessen it. 
3047- In June and July the sun rises to St. Helena in the south-east quadrant; 
about an hour after it passes into the north-east quadrant, and crosses it towards the 
southern end, being then at rnid-distance in the quadrant about one-third of the length, 
or nearly 60° from the southern termination. It leaves this quadrant about I'' 20“, 
crossing the magnetic meridian at that time (and consequently so long after passing 
the astronomical meridian), and entering the third or north-west quadrant traverses 
it obliquely towards its northern extremity. In our winter, December and January, 
the sun also rises to St. Helena in the south-east quadrant, as before; but it now 
remains in it until 22'', being for much of the time in strong places of action ; it 
enters the north-east quadrant to the south of St. Helena, and does not remain in it 
two hours, being then only in the weakest part of it ; it leaves it again before arriving 
at the astronomical meridian, then enters the north-west quadrant, gliding along near 
to its southern side, and is within two-thirds of an hour of leaving it when it sets to 
St. Helena. 
3048. As June presents the aspect of circumstances approaching nearest to that of 
a station further south, as Hobarton or the Cape, so I will consider the variations 
for it first. The north or upper end of the needle is then nearly at its mean place at 
midnight or 12'': it advances east (slowly at first) until 16'', and then more and more 
rapidly up to 19^ when it stops and goes as quickly west until about 22'', after which 
it changes but little until 3'', when it moves west till 5\ and then slowly east up to 
12*', and then onwards to 16'' and 19'', as already said. The eastening from midnight, 
and before I refer to the paramagnetic action of the cold, which comes up from the 
east as before (3003.3025.3037.) ; the rapid increase of the eastening from 16'' to 19'' 
is consistent with the increasing cold of the early morning, and also with the cir- 
cumstance, that the sun and its representative region are then passing from the south- 
east into the north-east quadrant, and must be not far from the neutral line, for that 
is the time of quickest transit of the needle. As the sun advances into the north-east 
quadrant, it first stops the eastening, as at 19'', and then converts it into westening 
(3014.), which goes on consistently with all former observations until 22'' ; the needle 
is then retained a little west of its mean position until H, at which time it has not 
yet attained coincidence with the magnetic meridian, and after this hour it is deter- 
mined east a little until 3''. This effect, from 22'' to 3'', I consider as the sun-swing 
to the east ; and I think, examining the globe (2906.), its small amount in declination 
is quite consistent with the relative positions of St. Helena and the warm region, 
combined with those of the active and neutral parts of the quadi’ants traversed during 
the time. From 3*“ to 5'' the needle end moves westward, following the sun ; and 
