112 DR. FARADAY’S EXPERIMENTAL RESEARCHES IN ELECTRICITY. (SERIES XXVII.) 
the effect of the atmosphere, though induced by the sun, lags behind the luminary, 
considered as in his astronomical position all the year round ; and that therefore in 
advancing to and receding from a tropic, he seems to do less in the first instance 
and more in the second than is due to his place for the time. 
3054. But where circumstances are apparently equal, a difference also arises. Thus 
from March to April in one direction, and from September to October in the other, 
might be expected to be alike, except for a little of the lagging effect (3053.), which 
would appear on both sides : nevertheless March and April are in Sabine’s curves 
between September and October, and near together, whilst the other two are far apart. 
Tliis effect I refer to the different conditions of the two hemispheres as regards heat 
(Dove). From September to October the sun is passing from a hemisphere having a 
mean temperature in summer 17°'4 above that of the other hemisphere for its winter; 
but in March and April it is departing from a hemisphere having a mean summer 
temperature of only 10°‘7 above that of the other hemisphere for its winter (2949.) ; 
and these respective differences must tend to separate September and October, and 
bring together March and April, as is seen to be the case by the curve charts, Plate II. 
3055. I need not go further into the declination variation of St. Helena: the lines 
for the other months are subject to the observations already made. Colonel Sabine’s 
important query of the cause of difference in direction for different months (3045.), 
appears to me at present to be answered for this station, as well as for the other 
stations, in very various latitudes where it makes its appearance (3016. 3022. 3039.). 
3056. The dip at St. Helena is a daily variation very simple in character, being a 
maximum at 7‘' and a minimum at 22*’ and 23*’, with only one progression. It pro- 
ceeds to its minimum therefore in the middle of the sun-swing, i. e. the upper end of 
the needle proceeds to west, and descends from lb** to 19*“ or 20*’, during which time 
therefore the dip is decreasing ; then it returns east until it reaches the neutral posi- 
tion, the dip decreasing the while still more. The needle still continues to go east to 
complete the sun-swing, but now the dip increases ; at 24** or P the needle returns 
(in declination) after the sun westward, but still with increasing dip; at 5*’ or 6** the 
westeninghas almost ceased, and an hour after the dip is at its maximum. 
3057. So as the sun and its region pass over they diminish the dip by depressing the 
upper ends of the lines of force, and as they pass away the lines rise (2926. 2937.) 
and the dip increases. The ellipse or curve, therefore, which represents the motion of 
the upper end of the needle at St. Helena, as the sun comes up from the east, is above 
westward and dowmward, and back below to east ; then rising to be repeated in the 
next twenty-four hours. This is the reverse direction to the representative ellipse for 
Hobarton, having like south dip in a greater degree. But that is in perfect consist- 
ency with the hypothesis ; for as the region is located above in the air, it is above the 
angle which the dip makes with the horizon at St. Helena, and therefore ought to 
depress the line of force and lessen the dip. At Hobarton, the region being in the 
tropical parts, is within the angle formed by the line of dip with the horizon, and 
